WHAT NEW IN X11.06 FRI

X 上的 AI 合集档案 · 23 个合集 · 69,408 条

gemini-4POST

实测评价:Gemini 4 复杂任务接近 Opus 5.5,安全任务意外强

用户称 Gemini 4 复杂任务接近甚至可能优于 Opus 5.5,安全类任务表现突出,整体远强于 Gemini 3.8 Flash,但仍不及 fable 5.1。

Haider.
@haider117 天前收录原推发布于 2026.10.18
Haider. @haider1原推发布于 2026.10.18
在 X 上打开原推

gemini 4 is definitely not on the same level as fable 5.1 but i'd put it close to opus 5.5 or maybe even better on complex tasks it's also surprisingly strong at security tasks even without access to the Fairwind (cybersecurity program) overall a much more capable model than gemini 3.8 flash

x.com/haider1/status/2111686209445769498采集于 2026.10.19 09:58