Anthropic, the company behind Claude, has just posted the kind of numbers most startups can only dream of. Quarterly revenue jumped from $11.5 billion to $16.8 billion, a 45% increase quarter over quarter.

And the good news doesn't stop there. According to adjusted calculations, the company says it has closed its last two quarters in profit. Every $1 spent on computing infrastructure is said to generate $1.70 in revenue. If forecasts hold, annualized revenue could cross the $110 billion mark before the end of the year.

A massive valuation target ahead of going public

Anthropic-IPO-xxxxx1 (1)

The big story, though, is the IPO. Anthropic reportedly plans to reach a valuation of more than $2 trillion before listing on the stock market. Part of the plan is to go public ahead of OpenAI — meaning the rivalry between the two AI heavyweights would play out not just in model development, but on the stock exchange as well.

That ambitious valuation comes with an equally ambitious price tag. Anthropic is expected to invest hundreds of billions of dollars in cloud and computing infrastructure over the coming decade. Training models and serving users demands serious compute, and expanding capacity as demand grows remains one of the company's biggest expense items.

Safety as the centerpiece of competition

Standing across from Anthropic are strong players like OpenAI — plus cheaper, China-based models that are muddying the picture. Simply building a powerful model isn't enough; running costs and winning user preference are critical too.

Anthropic differentiates itself through the weight it places on AI safety. Its growth strategy rests on the assumption that this approach can be pursued alongside a massive commercial operation. The revenue and profitability figures offer signals about the financial side of that strategy — but how the years ahead, with hundreds of billions in infrastructure spending coming into play, will transform today's picture remains an open question.