The last time grok made these statement, I tried using it for my workflows and it did not perform as good as opus or even sonnet. My guess is that xai benchmax

The last time grok made these statement, I tried using it for my workflows and it did not perform as good as opus or even sonnet. My guess is that xai benchmaxxes a lot but fails in actual capacity to produce good models.