deepseek-v4.1POST

日常使用 19.5 小时对比：GLM 5.3 Flash 177.2 tok/s 快于 DeepSeek V4.1 Fl

把 DeepSeek V4.1 Flash 当主力模型用了 19.5 小时后换回 GLM 5.3 Flash：同样 4 个框 8 个对话，GLM 177.2 tok/s、DeepSeek 140.9 tok/s，但 DeepSeek 的构建界面更好看。

CryptoBandit

@CryptoBandit10010 天前收录原推发布于 2026.10.22

CryptoBandit @CryptoBandit100原推发布于 2026.10.22

在 X 上打开原推

Ran DeepSeek V4.1 Flash as my daily model for 19.5 hours, then went back to GLM 5.3 Flash. Same 4 boxes, 8 chats: GLM 177.2 tok/s DeepSeek 140.9 tok/s GLM is faster. DeepSeek's builds look nicer. Which one would you run every day, and why? 👇

x.com/cryptobandit100/status/2113057753409138917采集于 2026.10.22 11:58