David

[David @David217820]

@Beimo_ai 很玄学，我一直正常用， 我是菲区20x，然后走洛杉矶ip

[David @David217820]

@Beimo_ai 朋友是不同的设备用了不同的机场节点， 他现在所有会话都显示Selected model is at capacity. Please try a different model.，甚至几百秒重试，这玩意就是玄学， 我也是好几台设备，都是网页，codex，不过都套了同一个isp

[David @David217820]

@thsottiaux Tibo, 速度什么可以提���来？当前fast也不快

[David @David217820]

@jlcjak 感谢��的测试，一开始我看到这个kicad的演示，我还准备规划一个新产品，但是现在还是等下一代吧。

[David @David217820]

@antigravity 感受到了，比3.7flash更快更准确，非常好用

[David @David217820]

@thsottiaux Reasoning is getting slower and slower in execution

[David @David217820]

@thsottiaux Will there be a model with a cliff-like improvement?

[David @David217820]

@DoWeMatchOrg @thsottiaux Tibo：you lost week limit

[David @David217820]

@thsottiaux Based on the performance of the ChatGPT web version's route to 5.5 mini, I am now highly skeptical of Codex as well and no longer dare to use it for real work.

[David @David217820]

@thsottiaux 5.5 mini

[David @David217820]

@OpenAI Why did you choose the O3 model, the 5.5 model, and the 5.6 SOL model? For each level of thinking depth, the reply was always the 5.5 mini model.

[David @David217820]

@wbs1231 感觉来回波动，时快时慢

[David @David217820]

@thsottiaux The latest version of ChatGPT desktop Windows will cause the mouse to get stuck.

[David @David217820]

@thsottiaux @hqmank it works

[David @David217820]

@thsottiaux Why is pro routed to 5.5-mini?

[David @David217820]

@OpenAI why 5.5 mini?

[David @David217820]

@Perp338 same.

[David @David217820]

@thsottiaux A very positive revision.

[David @David217820]

@thsottiaux The Codex Extension has become difficult to open.