实测者称 Gemini 4 Pro 明显弱于 Opus 5.5 与 Astra

Token Gremlin @TokenGremlin原推发布于 2026.09.30

在 X 上打开原推

From the tests I’ve been able to verify so far, Gemini 4 Pro looks pretty weak compared to Opus 5.5 and GPT-6 Astra. Honestly, I don’t even see much room for it to match either of them right now. The gap looks pretty brutal, and in some areas it seems closer to strong open models like GLM and Kimi than to the current frontier leaders. I really hope Google proves us wrong and ships something genuinely great.

这条是引用着另一条写的：「用户吐槽新 Gemini 4 Pro checkpoint 是浪费时间」