Elon Musk Calls Claude Opus 5.5 a 'Masterpiece', Admits Grok 4.8 Won't Match It
Summary
Elon Musk said Grok 4.8 is unlikely to match Anthropic’s Claude Opus 5.5, which he called a “masterpiece.” He expects Grok 4.8 to begin a limited rollout to some users later this week, after completing pretraining and continuing reinforcement learning. xAI has rewritten parts of its training and inference software in C and C++, but Musk gave no performance data. He predicts the next-generation Grok 5 could surpass Opus 5.5 in two or three months, without providing a firm release date. Anthropic reported a 66.4% score for Opus 5.5 on Terminal-Bench 4.0, compared with 57.9% for OpenAI’s GPT-6 Astra; xAI has published no comparable Grok 4.8 results.
