xAI just caught up (Grok 4.6 is here)

Developer Theo tests xAI's new Grok 4.6 model against Claude, GPT, and Kimi on coding, design, and long agent tasks, with mixed results.