质疑榜单：Gemini 3.7 Flash 分数高于 GPT-6 Astra

评论指出该 Intelligence Index 榜单里 Gemini 3.7 Flash 超过 GPT-6 Astra，且约一半任务是私有题。

Mert Gulsun

@mert_gulsun4 天前收录原推发布于 2026.10.24

Mert Gulsun @mert_gulsun原推发布于 2026.10.24

@BrendanFoody Isn't this your benchmark, with Gemini 3.7 Flash beating GPT-6 Astra? Also, ~1/2 of the Intelligence Index is private tasks.

x.com/mert_gulsun/status/2113785939667636511采集于 2026.10.24 10:08