Grok 4.7 gave me a pretty disappointing result in my 4 Elements benchmark.
You can obviously see a jump here, fire is much better, but keep in mind we now have Extra High, and the second result took 29 minutes to finish, while the first one took only 4 minutes and 26 seconds.
For this particular 3D / Three.js task, I wouldn't say Grok 4.7 is there yet.
I was really hoping we'd get Opus 5 level results, but I guess we'll have to wait a little bit longer for Grok in this use case.
I'll definitely test it more on long running tasks and complex stuff.
Follow along. I'll share everything here. 🔥