Elon Musk challenges Anthropic and OpenAI, says SpaceX will beat their AI in 6 months

Elon Musk has said SpaceX could match advanced AI models from OpenAI and Anthropic within months. The claim comes before a White House dinner where AI risks, regulation and competition are in focus.

In Short

• Musk said SpaceX could build a Fable or GPT-6 level model soon
• He said sustained momentum could put SpaceX in pole position within months
• The remarks came before a White House dinner with Trump and Xi

Anthropic and OpenAI are currently leading the AI race, with their latest models, Claude Fable 5.1 and GPT-6, setting the bar for what frontier AI can do. Elon Musk, however, thinks SpaceX could catch up sooner with its upcoming models. The billionaire claims that he is “cautiously optimistic” that SpaceX — which acquired xAI earlier this year — will have a Fable and GPT-6-level model in just two to three months, and could reach the top of the AI race within six months.

Musk’s claims about SpaceX’s future AI models came on X while responding to a discussion about Grok 4.7 and how xAI’s latest model compares with competing AI models from OpenAI and Anthropic. In response, Musk said he is optimistic about SpaceX’s progress. “I am cautiously optimistic that SpaceX will have a Fable/GPT-6 level model in 2 to 3 months,” he wrote.

The SpaceX CEO then responded to another X user who suggested that OpenAI and Anthropic could simply release more advanced models by the time SpaceX catches up. Musk said SpaceX would continue accelerating its AI efforts and claimed that, if its pace of improvement remains strong, the company could reach the top of the AI field within about six months.

“We will keep accelerating. Our AI efforts are only 3 years old, vs 6 and 10 years old for Anthropic and OpenAI. If our second derivative remains strong, SpaceX will reach pole position in about 6 months,” Musk wrote.

As for why he is confident, Musk highlighted AI capability and computing infrastructure as two factors. He argued that once an AI model is already more intelligent than necessary for a particular class of tasks, adding even more intelligence may not offer much practical benefit. “Once you far exceed the caliber of intelligence needed for a class of tasks, additional intelligence is pointless. You don’t need (and it would be cruel to put) Newton-level intelligence in your toaster,” he wrote.

He also pointed to the difficulty of bringing massive amounts of computing power online quickly, saying SpaceX has demonstrated an ability to do this and will continue to improve. “Hardware is hard. Bringing massive compute online rapidly is incredibly difficult. SpaceX has demonstrated exceptional ability in this regard and will only get better,” he added.

Essentially, Musk is saying that his company SpaceX is accelerating its AI development, and if that progress continues, it could catch up with OpenAI and Anthropic within months and potentially move ahead of them in the AI race.

Musk’s comments come at a time when both Anthropic and OpenAI are pushing the capabilities of their latest AI models. Anthropic says its latest model, Claude Fable 5.1, is its most powerful model yet, and is designed for demanding coding, knowledge work, long-running problem-solving and research tasks. It also comes with a 1-million-token context window.

OpenAI, meanwhile, is also touting the capabilities of its latest model, GPT-6 Astra, which it describes as its most capable model yet from its lab. The company says Astra performs strongly in areas such as computer use, software engineering, science, cybersecurity and other professional tasks. It can also handle complex, multi-step work, including creating documents and spreadsheets, conducting research and working with software.

- Ends