Event ·Multi-source

GPT-5.6 Sol Pro几乎饱和prinzbench基准

GPT-5.6 Sol Pro在prinzbench上取得91/99的成绩，基本饱和该基准。该基准包含两道从未被任何模型解出的问题（各值3分），排除后模型正确回答了93题中的91题。相比GPT-5.4 Pro的79/99和GPT-5.5 Pro的82/99有明显提升。prinzbench于2026年1月发布，仅5个月后被GPT-5.6 Sol Pro饱和，加速趋势显著。

Current read

GPT-5.6 Sol Pro把prinzbench刷到91分，比前代高出一截，连从未解出的题都快能答了，基准加速饱和太明显。

11 sources48° AI heatLast updated 2026/7/16 22:46:12

Evidence

11 sources

Primary sourceGreg Brockman

2026/7/16 22:46:12View source

Related source 1Decoder

2026/7/15 17:35:12View source

Related source 2arXiv: OpenAI

2026/7/15 19:08:56View source

Related source 3IT之家

2026/7/16 02:51:14View source

Related source 4OpenAI

2026/7/17 21:00:01View source

Related source 5elvis

2026/7/15 17:51:11View source

Related source 6shao__meng

2026/7/16 02:24:07View source

Related source 7Suhail

2026/7/16 16:11:45View source

Related source 8marktechpost

2026/7/16 18:48:04View source

Related source 9@OpenAIDevs

2026/7/17 21:01:18View source

Related source 10歸藏(guizang.ai)

2026/7/18 02:18:56View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story