Elon Musk Admits Anthropic Has Done ‘Something Special With Coding’ Despite xAI Being Blocked

Elon Musk, the founder and CEO of xAI, said in a post on X that the company’s upcoming model, Grok 4.2, will be better than Anthropic’s Claude Opus 4.5 in several aspects — but “not quite in programming.”

A few days ago, Musk responded to a user on X who praised Claude Opus 4.5’s benchmark performance, saying that Grok “might do better” in its next iteration, Grok 4.2. He later clarified that while Grok 4.2 could surpass Claude in several areas, coding would not be one of them.

“Anthropic has done something special with coding,” Musk said. 

Surprisingly, Musk’s praise for Anthropic comes at a time when the company has blocked xAI’s access to Claude models. 

Recently, Anthropic restricted the use of its AI models by rival labs, including xAI. According to journalist Kylie Robinson’s post on X, xAI staff used Anthropic’s AI models internally via the Cursor platform until Anthropic cut off access. 

Robinson cited an internal message she viewed from xAI co-founder Tony Wu sent to the team, which revealed that, “I believe many of you have already discovered that anthropic models are not responding on cursor. According to Cursor, this is a new policy anthropic is enforcing for all its major competitors.”

“This is both bad and good news. We will get a hit on productivity, but it [really] pushes us to develop our own coding product/models,” Wu appeared to add in the memo. 

Currently, Anthropic’s Claude Opus 4.5 model leads several benchmarks on coding evaluations, outperforming xAI’s Grok 4 and Grok 4.1 models, and even the company’s coding-focused model, Grok Code. 

That said, Musk stated last week that the Grok Code is set to receive a major upgrade next month. “It will one-shot many complex coding tasks.”