HypoArena - Biomedical Science: leaderboard

Metric: Rating on the 244 biomedical science cases: Bradley-Terry-Davidson rating (base 1500, open-ended) from reference-independent pairwise judgments by seed-2.0-pro, both presentation orders, with the source-derived reference hypothesis set as an anonymous competitor; baseline single-pass generation; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 15 models tracked.

Top models

#ModelScore
1GPT-5.4 (High)1627.2
2Claude Opus 4.6 (High)1608.9
3Claude Sonnet 4.6 (High)1601.2
4GLM-5.11592.8
5Kimi K2.61570.5
6DeepSeek V4 Pro (High)1525
7MiniMax-M2.71508.3
8GLM-51469.1
9DeepSeek V4 Flash (High)1464.3
10MiniMax-M2.51458.4
11Gemini 3.1 Pro (Preview) (High)1429.9
12GPT-5.4 Mini (High)1424
13Gemini 3 Flash (High)1420
14Kimi K2.5 (Thinking)1281.8

Interactive version: theaggregate.ai/benchmark?slug=hypoarena-biomedical-science · How It Works · Data refreshed daily, snapshot 2026-09-29.