Rows of servers inside a data centre, as Anthropic's new filing reveals $17 billion in quarterly revenue and $895 billion in spending commitments.

Anthropic’s business is growing at a pace that would be extraordinary for any company — and its spending plans are larger still, according to a new filing that lays bare both sides of the AI boom.

Key Points

Server racks in a data centre, the computing infrastructure behind Anthropic's $895 billion in spending commitments.
  • Anthropic reported third-quarter revenue of $17 billion, up from $1.4 billion a year earlier — a rise of more than 1,100% — in an amended filing reported on Monday 2 November.
  • The net loss for July–September reached $23.3 billion, while the operating loss was $567 million, as the company paid more for short-term computing capacity.
  • The company has committed to spend about $895 billion, mostly on computing power, with $70 billion due before September 2027.
  • Reports put its planned stock-market debut at raising about $100 billion at a valuation of $1.5–$2 trillion, later this month.

Growth, and the bill that comes with it

The headline figure is startling: revenue of $17 billion in a single quarter, against $1.4 billion in the same period a year earlier. Few companies in history have grown that fast at that scale, and the filing underlines why Anthropic’s planned initial public offering is being treated as a bellwether for the entire AI industry.

But the filing also shows the cost of that growth. A net loss of $23.3 billion in the quarter, and an operating loss of $567 million driven by payments for computing capacity, illustrate the central tension of the AI business: the models that generate the revenue can only run on vast, expensive infrastructure, and that infrastructure must be bought before the revenue arrives.

An $895 billion promise

The most eye-catching number in the filing is not revenue but commitment: about $895 billion pledged, mostly for computing power, with $70 billion of it due before September 2027. It is a bet that demand for AI will keep growing fast enough to justify spending on a scale normally associated with national budgets.

That bet is also the main risk. The filing itself runs to around 400 pages, with more than 90 devoted to risks — including AI systems behaving unexpectedly — and sales growth has already slowed from previous quarters as competition intensifies from OpenAI and from cheaper Chinese models. Investors weighing the IPO must decide whether Anthropic is buying the future at a fair price, or mortgaging it.

What happens next

If the offering goes ahead later this month at anything near the reported $1.5–$2 trillion valuation, it will be one of the largest listings in history — and a public test of how long the AI spending boom can last. Tuesday’s filing gives the market its clearest picture yet of what it would be buying: extraordinary growth, extraordinary losses, and an extraordinary bill still to come.

Related coverage: planned initial public offering · Technology · AI build-out

Sources

Leave a Reply

Your email address will not be published. Required fields are marked *