Have you tried SuperHeavy yet? It seems to be crushing on almost any measurable ...

Rover222 on July 26, 2025 | parent | context | favorite | on: All AI models might be the same

Have you tried SuperHeavy yet? It seems to be crushing on almost any measurable benchmark. I've only used Grok 4 (not heavy) when stuck on a problem that claude/cursor cannot get past, and Grok 4 is often able to come up with an elegant solution.