Explore / App / JevArena (Blind Decision Evaluation)

JevArena (Blind Decision Evaluation)

Self-hostWebTerminal

Compare judgments before revealing model identities and costs.

by admin · listed 7 days ago

⚡ What System1 does here

Sends the same candidate question to Jev and another model judge, presenting structured answers for comparison.

About

A web tool for blind comparison: users supply a question and candidate answers, compare judgments from Jev and another model, then vote to reveal identities, latency and cost. The public site requires the user's own OpenRouter key. A research CLI is also available. Preference votes and model accuracy are explicitly treated as different measures. Source code is available free of charge. Online models, third-party services and related software may incur additional costs. Sources: https://github.com/chenmingtang830/jevarena/blob/main/README.md https://www.jev.store/projects/jevarena-lab This review checked source code and documentation only. The application was not installed or run, and its practical effectiveness was not verified.

🧪 Community test reports

0 reports · 0 passed on the real model

No one has tested this yet. Be the first.

💬 Discussions

New post →

No discussions yet.