This BioMysteryBench subset contains the 23 tasks that no member of the human-expert panel solved correctly. Tasks still have controlled objective answers and use real-world bioinformatics data, tools and restricted external data access.
查看 BioMysteryBench (Human-difficult) 的最新得分、模型模式、发布时间与参数规模，快速了解当前完整榜单表现。
数据优先来自官方发布（GitHub、Hugging Face、论文），其次为评测基准官方结果，最后为第三方评测机构数据。 了解数据收集方法
| 排名 | 模型 | 开源情况 | |||
|---|---|---|---|---|---|
 Gemini 3.7 Flash 思考水平·中工具联网 | 43.50 | 2026-08-13 | 未知 | 闭源 |
Google reported 43.5% for Gemini 3.7 Flash on the BioMysteryBench human-difficult subset in August 2026.
