Anthropic signals its first profit – and overtakes OpenAI in value

Anthropic has told its investors it expects to post its first-ever operating profit in the current quarter. The figures shared with investors point to roughly $10.9 billion in revenue for the second quarter – the one ending in June – up from $4.8 billion in the first. That is 130 percent growth in three months, and the bottom line flips from red to an operating profit estimated at $559 million.

Less than a year ago the same company told investors it did not expect full-year profitability before 2028 at the earliest. Now it is projecting a profit in a single quarter. CEO Dario Amodei has described the growth as almost too hard to handle, pointing to demand for coding tools as the main driver. Claude has become something developers pay for at work, not an experiment.

The numbers carry important caveats

The profit deserves a sober reading. Anthropic itself warns that profitability may not hold for the full year, because it plans to sharply increase spending on compute and infrastructure. The profit figure includes the cost of training new frontier models, but it excludes stock-based compensation – a large line item at a company that pays staff in equity. Include it, and the picture looks different.

The direction, however, is real. The company says it spent 71 cents on compute per dollar of revenue in the first quarter and expects that to fall to 56 cents in the second. Margins improve because revenue is growing faster than the cost to serve each customer. That is the opposite of the story often told about AI companies: that they only burn cash.

Past OpenAI in value

In parallel, Anthropic is negotiating a funding round of more than $30 billion at a pre-money valuation above $900 billion. According to Bloomberg, Dragoneer, Greenoaks, Sequoia Capital and Altimeter Capital are lined up as lead investors, and the round is expected to close as soon as this week. Google and Amazon are already heavy strategic backers.

If the round closes as planned, Anthropic passes OpenAI, last valued at $852 billion in a March round. Two companies recently cast as challenger and leader now stand shoulder to shoulder at the top – one heading toward a public listing through a confidential filing, the other with an operating profit to show.

What leaders should take from this

For CIOs, CFOs and boards, the point is not who is biggest this month. The point is that the two dominant suppliers of generative AI now have the capital and earnings to cement their positions for years. That lowers the risk that one of them disappears, but raises the risk of vendor lock-in. When Claude and ChatGPT are both financially solid, it becomes easier to build critical workflows on them – and harder to switch when pricing, terms or data policy change.

The practical advice is the same as before, but it now carries more weight: treat model access as a procurement decision, not a technical detail. Negotiate exit terms and data portability before locking workflows to one model. Demand clarity on where data is processed and stored. And consider an architecture that lets you move between Anthropic, OpenAI and open alternatives without a rewrite. A profitable Anthropic is a more predictable supplier – but a stronger supplier negotiates harder.

The second signal is for finance. When even the most expensive link in the AI value chain starts to show a positive margin, it gets harder to defend internal AI projects that still cannot measure value. The question the board should ask is not whether AI pays off for Anthropic, but whether it pays off for you.

Sources and media

Primary source: CNBC, «Anthropic set to hit $10.9 billion in revenue in Q2, source says», 20 May 2026, source_url: https://www.cnbc.com/2026/05/20/anthropic-revenue-explosive-growth-ipo-profitable-quarter.html. Funding round and valuation figures reported by Bloomberg. Credit: CNBC and Bloomberg. Thumbnail generated with OpenAI Image 2 / hogby.ai (editorial illustration, no logos).