Grok 4.5 vs. Claude Opus: xAI's AI Progress Explained

Elon Musk claims xAI's Grok is advancing faster than any competitor. This statement follows the release of Grok 4.5, which became publicly available five days prior. What's interesting is Grok 4.5 is described as an "Opus-class model," meaning it's comparable to Anthropic's Claude Opus. Grok 4.5 scored 29.0% on the SWE Marathon benchmark, surpassing Opus 4.8's 26.0%. xAI also states Grok 4.5 runs at 80 transactions per second and has about twice the token efficiency of other leading models. Its pricing is $2 per million input tokens and $6 per million output tokens. The development pace is notable, with xAI reportedly pushing model updates twice a week. Grok Build received an update on the same day as Musk's comment, and Grok Imagine was declared complete just days before. The next milestone is Grok 4.5's availability in the European Union, targeted for mid-July. This will test xAI's operational maturity. This rapid progress and specific benchmark claims suggest significant movement in the AI landscape.