Fahad

I’m running Fable, K3, 5.6 and 4.5 like 10 hours a day and they are all sick. At work, at home. I have subscriptions for many of them at both locations. But you know which model legit doesn’t get enough hype? GROK. I’m so serious. Even on 𝕏 it doesn’t get enough hype cause we assume oh yeah it’s Elon’s network of course they are gonna hype grok. But I don’t care about benchmarks. It absolutely crushes in usefulness and im quickly using it more and more. It gets the job done well, and fast, and it doesn't spend all day thinking about nothing. But its not just fast and cheap. Sometimes it WORKS BETTER than 5.6 or even Fable. Check this. I told Grok and Codex to help design a level that I could dynamically spawn some quest monsters and players on using the level tools available for my game. Sol on max settings took forever to put out garbage and Grok's level was pretty sick right away. I have many other dramatic cases of Grok outperforming so called frontier models on genuinely useful tasks. I can’t wait for the 2 parameter model.