Anthropic's revenue surge and the race to IPO first

A year ago, Anthropic was still widely described as the scrappy, safety-focused underdog trailing OpenAI in the AI race. That description does not really hold up anymore. By the middle of 2026, Anthropic's annualized revenue run rate had rocketed from about 1 billion dollars in January 2025 to somewhere around 65 billion dollars, and the company is now on track to list on public markets before its much larger, much more famous rival does.

If that sounds like a plot twist, it is. It's also a reminder of just how fast the ground can shift underneath a rivalry that felt fairly settled as recently as early 2025. So let's dig into exactly how fast Anthropic has grown, where the money is actually coming from, and why some very serious investors think this might be the strongest evidence yet that the AI boom is not entirely a bubble.

How fast is Anthropic actually growing

Anthropic doesn't file public earnings yet, but the company shares regular updates with its investors, and enough of those figures have leaked out to trace an almost unbelievable curve. Its annualized revenue run rate stood at roughly 1 billion dollars in January 2025. By the end of that year it had reached about 9 billion dollars. It kept climbing to around 14 billion dollars in February 2026, then jumped past OpenAI's own run rate for the first time in April 2026 at about 30 billion dollars, based on leaked figures detailed in Sovereign Magazine's analysis.

The acceleration lines up closely with a string of enterprise-focused model releases through 2025 and 2026, along with a growing base of large companies moving past pilot programs and into production deployments of Anthropic's models across coding, customer support, and internal research tools. Growth like this rarely comes from one single product decision, it tends to come from a company hitting the right combination of product quality and timing right as enterprise AI budgets are expanding fastest.

The pace only accelerated from there. Anthropic disclosed a run rate near 47 billion dollars in May 2026, then told investors it had hit roughly 65 billion dollars by the end of July, a jump of about 38 percent in just two months. Quarterly numbers tell the same story from a different angle. Anthropic reportedly generated 4.73 billion dollars in the first quarter of 2026, then more than 11.5 billion dollars in the second quarter, up from just 787 million dollars in the same quarter a year earlier, according to detailed reporting from The Edge Malaysia. That's roughly a fourteenfold increase year over year in a single quarter.

Forecasts reportedly shared with investors put Anthropic's full 2026 revenue somewhere between 100 billion and 120 billion dollars, with a longer range target of 190 billion to 200 billion dollars by 2028.

Where Anthropic's money actually comes from

Unlike OpenAI, which built its business on top of hundreds of millions of ChatGPT users paying 20 dollars a month, Anthropic's growth has been driven overwhelmingly by developers and enterprises. The bulk of its revenue, by most industry estimates, comes from API usage and business contracts rather than consumer subscriptions to Claude.ai directly.

Coding has been the single biggest driver. AI coding assistants built on Anthropic's models, especially its own Claude Code product, have become genuinely popular with professional software teams, and Claude Code alone reportedly generates about 2.5 billion dollars in annualized revenue on its own. That kind of enterprise AI adoption tends to come with larger, stickier contracts than consumer subscriptions, which helps explain how Anthropic's revenue scaled so quickly without a ChatGPT-sized consumer app to lean on.

The valuation rocket ride

The fundraising side of this story has moved almost as fast as the revenue.

• March 2025, Series E, 3.5 billion dollars raised, valuing the company at 61.5 billion dollars
• September 2025, Series F, 13 billion dollars raised, valuing the company at 183 billion dollars
• February 2026, Series G, 30 billion dollars raised, valuing the company at 380 billion dollars, led by Coatue and GIC
• May 2026, Series H, 65 billion dollars raised, valuing the company at 965 billion dollars, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital

That May 2026 round briefly made Anthropic more valuable on paper than OpenAI, whose own valuation stood at 852 billion dollars after its March 2026 funding round. It's a remarkable reversal for a company that was valued at roughly 1.5 billion dollars back in 2022.

Why Anthropic might beat OpenAI to wall street

On June 1, 2026, Anthropic confidentially filed a draft registration statement with the Securities and Exchange Commission, the standard first step toward an initial public offering, as Anthropic's own announcement and TechCrunch's coverage both confirmed at the time. Investment banks Goldman Sachs, JPMorgan, and Morgan Stanley are reportedly running the process, and people close to the deal have said Anthropic could list as early as September or October 2026, well ahead of OpenAI's own target of sometime in 2027.

The math investors are running is genuinely interesting. Anthropic's 965 billion dollar valuation works out to roughly 15 times its most recent annualized revenue run rate, which actually looks conservative next to other high-growth companies heading toward public markets around the same time. Palantir has traded at around 53 times its expected 2026 revenue, while SpaceX and Cloudflare have both traded closer to 41.6 times. If Anthropic's revenue keeps compounding anywhere near its current pace, some bankers argue the eventual IPO could price well above that 965 billion dollar mark.

How Anthropic stacks up against the rest of the field

Anthropic and OpenAI aren't the only companies spending heavily to win this market, they're just the two furthest ahead by revenue. Google's Gemini has reportedly been gaining ground quickly on the consumer side, with its share of AI web traffic climbing from under 6 percent to more than 21 percent in about a year, largely at OpenAI's expense rather than Anthropic's. Meta and xAI continue to pour billions into their own models, but neither has disclosed revenue anywhere close to Anthropic's or OpenAI's scale.

That leaves Anthropic and OpenAI as the two companies genuinely racing each other toward public markets, and increasingly toward the same enterprise customers. Both are now structured as public benefit corporations rather than conventional for-profit companies, a structure both argue lets them balance commercial growth against their stated safety missions, though critics on both sides continue to debate whether that structure meaningfully changes day to day decisions once billions of dollars and a looming IPO are on the table.

The profit question that actually sets Anthropic apart

Here's where the story gets genuinely different from OpenAI's. Anthropic reported positive adjusted operating income for the second quarter of 2026, its first quarter in the black on that basis, even while spending an estimated 19 billion dollars a year on computing. Gross margin has reportedly climbed to around 40 percent, with the company telling investors it's targeting something closer to 77 percent by 2028 as it scales. Part of that margin improvement appears to come from pricing power, enterprise and API contracts tend to carry higher effective per-token pricing than consumer subscriptions, and part of it comes from ordinary efficiency gains that show up once a company is running models at this kind of scale.

One good quarter of positive adjusted operating income does not mean Anthropic has fully solved the economics of frontier AI, and a single quarter is not the same thing as a durable trend. But it's a meaningfully different financial position than OpenAI, which has told its own investors to expect tens of billions of dollars in annual losses well into the end of the decade. For an industry that's been repeatedly accused of burning cash with no real path to profitability, Anthropic's numbers are about the closest thing the AI industry has to a counterargument right now.

The risks nobody should ignore

None of this makes Anthropic a sure thing. In June 2026, the company had to temporarily suspend access to two of its most advanced models, Claude Fable 5 and Mythos 5, after the U.S. Department of Commerce issued export controls citing national security authorities. Anthropic restored access about two weeks later once the restrictions were lifted, and Anthropic's own statement on the episode remains the clearest account of what happened, but the incident was a pointed reminder that Anthropic's business now sits squarely inside U.S. trade and national security policy, not just the technology market.

Investors weighing the IPO have also raised questions about lower-cost competing models eating into margins over time, along with local opposition to new data center construction and a relationship with the current administration that reporting has described as increasingly strained. Anthropic, like OpenAI, also depends on a fairly small number of cloud and chip partners for the computing power behind its growth, a concentration risk that shows up across nearly every company racing to build frontier AI right now. None of this is unique to Anthropic, its competitors face similar pressure, but it's worth keeping in mind before treating a 965 billion dollar valuation as a settled number.

What this means for the broader AI race

Step back far enough and the Anthropic story looks like a genuine changing of the guard, at least on revenue. OpenAI still has the bigger consumer brand and hundreds of millions more monthly users, but Anthropic's foundation models have won over enterprise buyers and developers to the point that its revenue run rate pulled ahead of OpenAI's sometime around April 2026 and has stayed there since.

Whether that lead holds once an actual IPO prospectus puts audited numbers in front of the public, and whether either company can turn today's astonishing growth rates into durable, sustainable profit, is still very much an open question. But for anyone trying to understand where the real money in AI is heading, watching whether Anthropic actually beats OpenAI to the public markets this fall is about as good a signal as you're going to get.