The generative-AI race has a new revenue leader, and for the first time it is not the company that made "ChatGPT" a household word. Anthropic, the San Francisco lab founded by former OpenAI researchers, has overtaken OpenAI in self-reported revenue and is now running at a roughly $47 billion annualized clip — a figure it says will make it profitable in 2026, a full year ahead of its own prior guidance. Fortune confirmed the crossover on July 2, and the underlying data suggests it is not a fluke of timing but a structural shift in where enterprise dollars are flowing.
The scale of the reversal is hard to overstate. As recently as last year, Anthropic was the perennial number two, trailing OpenAI on revenue, users, and mindshare. In May it told investors it was on course to hit a $47 billion revenue run rate this year, up from a $30 billion run rate earlier in 2026 and roughly $10 billion in annual revenue in 2025. OpenAI, by contrast, has guided to $25 billion to $33 billion in annualized revenue for 2026 while projecting an operating loss of about $14 billion — losses that internal documents peg at nearly three times the prior year's. OpenAI still dwarfs its rival in reach, with roughly 1.1 billion monthly active users to Claude's far smaller consumer base. But on the metric that matters most to the IPO bankers now circling both companies, Anthropic has pulled ahead.
How the enterprise flipped
The engine of the shift is code. Claude Code, Anthropic's agentic software-development tool, launched publicly in May 2025 and hit roughly $1 billion in annualized run-rate revenue by the end of the year. In the three months that followed, it more than doubled to about $2.5 billion by February 2026 — one of the fastest revenue ramps of any software product on record. Anthropic says enterprise use now accounts for more than half of Claude Code's revenue, and that business subscriptions across its products have quadrupled since the start of the year.
That enterprise momentum shows up in independent data. Corporate-spend figures from Ramp show Anthropic overtook OpenAI on business subscriptions in May 2026, a signal that finance and engineering teams — not just individual hobbyists — are choosing Claude. On the consumer side, the picture is inverting for OpenAI: Similarweb data shows monthly visits to ChatGPT fell below a majority of the generative-AI market for the first time in May, the first month the product no longer commanded the plurality of consumer traffic it had held since launch.
"Consumers are increasingly willing to switch between models," noted Deutsche Bank Research Institute analyst Adrian Cox, who has tracked the widening competitive field. The loyalty that once made ChatGPT synonymous with AI, in other words, is proving thinner than assumed — and Anthropic has exploited the opening by going where the switching costs are highest and the budgets are largest.
Why it matters
The revenue crossover reframes a rivalry that has defined the AI era. Anthropic completed a $65 billion Series H in late May that valued it at roughly $965 billion, edging past OpenAI to become the most valuable AI startup and putting it within striking distance of a trillion-dollar valuation. It is targeting an initial public offering as soon as October 2026, and hitting profitability this year would give it something OpenAI conspicuously lacks heading into public markets: a credible path to positive operating margins.
The contrast is stark. OpenAI is spending aggressively on compute and consumer scale and, by its own forecast, will burn about $14 billion this year even as it seeks tens of billions more in funding. Anthropic has instead concentrated on high-margin enterprise and developer products, and the bet is paying off in the numbers that IPO investors scrutinize first. For a sector where nearly every leading player runs at a loss, a profitable AI lab racing toward a public listing is a genuinely new proposition.
There is a risk in the euphoria, and Cox has flagged it. "They and their peers need to strike while the iron is hot to secure computing power, distribution and capital," he wrote, warning that if investors lose patience with AI's revenue prospects, access to capital could shrink quickly. The window for both labs to convert momentum into durable advantage may be narrower than the current fundraising frenzy implies.
What to watch next
The next few months will test whether Anthropic's lead is a lasting reordering or a snapshot. Watch the October IPO timeline: any slip would signal caution about public-market appetite, while a strong debut would validate the enterprise-first model over OpenAI's growth-at-any-cost approach. Watch, too, whether OpenAI responds by leaning harder into monetization — ads in ChatGPT and new enterprise tiers are already on the table — and whether Claude Code's blistering growth rate holds or normalizes as competitors ship rival coding agents. The revenue crown has changed hands once this year. In a market moving this fast, it may not stay put.
"They and their peers need to strike while the iron is hot to secure computing power, distribution and capital."— Adrian Cox, Analyst, Deutsche Bank Research Institute