Dogan Ural
@doganuraldesign · 18. Nov. 2025
If this is true, Gemini 3 is cooked.
I made a quick table with Grok 4 and Grok 4.1 (estimated) scores.
Now you know why they didn't include Grok.
Grok 4 scored either superior or equivalent results in most benchmarks. And the Grok 4.1 results aren't even out yet.
Elon Musk
@elonmusk
Hmm
14:59 · 18. November 2025 · 79.061 Aufrufe
129
40
608