Gemini 4.0 Pro 实测：担心被 GPT-6 Sol 抢风头

Hakm @hakmgpt原推发布于 2026.09.18

在 X 上打开原推

After testing Gemini 4.0 Pro on Arena, I can share my honest thoughts on the model. In some areas, it can feel same level of 5.6 Sol and Opus 5. It outperform Muse Spark 1.3 in my tests , it’s great with 3D and SVG However, to be completely honest, I think it could end up in a similar position to 3.8 Flash today it’s good model, but the problem is the timing. With upcoming models like Claude Opus 5.2, GPT-6 Sol, and Fable 5.2, it may struggle to remain competitive in the long run.