Anthropic, the San Francisco-based artificial intelligence startup behind the Claude AI assistant, is on the verge of recording its first quarterly operating profit, even as it commits to one of the largest computing expenditures in the industry's history. The company's June quarter revenue is expected to reach at least $10.9 billion — more than double the $4.8 billion it recorded in the March quarter — with an anticipated operating profit of $559 million, according to a person familiar with the matter who spoke to Reuters on condition of anonymity. The figures were first reported by The Wall Street Journal.

Driving the surge in revenue is soaring demand for Claude, Anthropic's flagship AI model. Software developers are increasingly relying on it to assist with computer programming, while enterprises are deploying the company's top-tier model, Mythos, to identify vulnerabilities in their code. The milestone is a notable one in an industry where profitability has remained elusive, as the steep costs of developing and running AI systems have kept most major players firmly in the red.

Those costs were thrown into sharp relief on Wednesday by a separate disclosure. SpaceX — the rocket and satellite company founded by Elon Musk — revealed in its IPO filing that Anthropic had agreed to pay it $1.25 billion per month for computing capacity through May 2029. The deal covers access to both of SpaceX's AI training data centre clusters, known as Colossus and Colossus I, with reduced fees during a capacity ramp-up period this month and next. Either party may exit the arrangement with 90 days' notice.

For SpaceX, the agreement represents a significant commercial lifeline for its AI segment, which reported an operating loss of approximately $2.5 billion in the March quarter against segment revenue of $818 million. Musk indicated on his social media platform X that SpaceX is in discussions with additional companies about offering AI computing capacity as a service at scale, signalling an ambition to become a major player in the cloud infrastructure market.

The developments highlight a defining tension in the AI industry: even as companies like Anthropic approach financial sustainability, the infrastructure required to stay competitive demands expenditure of almost inconceivable scale. That Anthropic can now plausibly absorb a $1.25 billion monthly computing bill while still projecting a profit underscores just how rapidly the economics of commercial AI are shifting.