返回
原文
Sherwin Wu@sherwinwu
43AI 编辑部评分，满分 100
2026-08-14 01:35· 20分钟前
AI 导读

Cerebras 预览 GPT-5.6 Sol Ultrafast 模式，生成速度最高 750 tokens/秒，比标准处理快 14 倍。该模式跑完 Humanity's Last Exam 仅用 11 小时 11 分，比 Claude Fable 5 快近 7 倍且准确率相当。用户称瓶颈已从模型转向工具本身。

GPT-5.6 Sol Ultrafast is here!!

750 tokens per second honestly feels instantaneous for most workloads. The bottleneck now moves to the tools themselves.

CerebrasPreviewing Ultrafast mode for @OpenAI's GPT 5.6 Sol, powered by Cerebras. GPT-5.6 Sol Ultrafast generates responses at up to 750 tokens per second. That's the f...
OpenAI产品更新推理

来源：Sherwin Wu · x.com

导出 Markdown
返回
AI 评分 43/100在 X 查看原推
导出 Markdown
Sherwin Wu · @sherwinwu · X·2026-08-14 01:35·20分钟前
AI 导读

Cerebras 预览 GPT-5.6 Sol Ultrafast 模式，生成速度最高 750 tokens/秒，比标准处理快 14 倍。该模式跑完 Humanity's Last Exam 仅用 11 小时 11 分，比 Claude Fable 5 快近 7 倍且准确率相当。用户称瓶颈已从模型转向工具本身。

GPT-5.6 Sol Ultrafast is here!!

750 tokens per second honestly feels instantaneous for most workloads. The bottleneck now moves to the tools themselves.

CerebrasPreviewing Ultrafast mode for @OpenAI's GPT 5.6 Sol, powered by Cerebras. GPT-5.6 Sol Ultrafast generates responses at up to 750 tokens per second. That's the f...

来源：Sherwin Wu· x.com

OpenAI产品更新推理