How It Works

Learn how your votes power transparent AI progress and help shape the future of artificial intelligence

The Process

Four simple steps to help improve AI for everyone

1

Ask a Question

Type in your prompt or question to our AI comparison arena. Your data helps us build fair, transparent AI evaluations.

Submit any question or prompt you'd like to test. Our system anonymously serves your prompt to multiple AI models simultaneously.

2

Compare Responses

You'll see responses from 2 anonymous AI models. Analyze the answers and decide which one better fits your needs.

Models are completely anonymous during evaluation to prevent bias. Focus purely on response quality and helpfulness.

3

Vote for the Best

Choose your preferred response. Your vote directly impacts our public AI leaderboards and helps shape AI development.

Every vote matters. Your preferences help determine which models perform best for real-world use cases.

4

Discover and Repeat

After voting, model identities are revealed. Keep exploring to help improve AI in our community-powered evaluation space.

See which models you preferred and continue testing to build comprehensive AI performance data.

Platform Features

Comprehensive tools for AI evaluation and research

Voice Conversations

Participate in real-time voice conversations with AI models and provide feedback on natural interaction quality.

Anonymous Evaluation

All model comparisons are completely anonymous to ensure unbiased evaluation based purely on response quality.

Public Leaderboards

Your votes directly shape our transparent, community-driven AI model rankings that anyone can access.

Open Research

We release datasets and research findings to help the broader AI community understand human preferences.

Transparency & Privacy

Learn more about our policies and how the community impacts the leaderboards

Privacy First

Your data is protected and anonymized. We only collect what's necessary for fair evaluation.

Open Platform

Our leaderboards are public and transparent. Anyone can see how models perform.

Community Driven

Every vote counts. The community directly shapes which models rank highest.

Since our launch, we've helped evaluate proprietary and open source models from major labs and innovative teams. This includes pre-release models, meaning the community's feedback directly influences how new AI models are developed, refined, and released.

Ready to Shape AI's Future?

Join thousands of users helping evaluate and improve AI models through transparent, community-driven research.