Anthropic just did something that would have sounded like fan fiction a year ago: on secondary markets, its implied valuation has ripped past the $1 trillion mark, leapfrogging OpenAI and turning the AI lab into the hottest private stock in Silicon Valley.

The wild part is that this isn’t coming from a new primary funding round or an IPO filing. Anthropic’s last official valuation was $380 billion in a Series G round in February 2026, when it raised $30 billion led by GIC and Coatue, more than doubling its earlier $183 billion Series F price from September 2025. On paper, that already made the Claude maker one of the most richly valued startups in history. But in the shadowy world of secondary markets, where existing shareholders quietly unload slivers of equity to latecomers desperate for a piece of the action, prices have detached from those funding-round numbers in a big way.

On Forge Global, one of the biggest private stock marketplaces, Anthropic’s “Forge Price” has surged so high that it implies an overall valuation of roughly $1 trillion, according to the company’s CEO, who spoke to Business Insider. That would put Anthropic above OpenAI, whose secondary valuation on the same platform sits closer to $880 billion, only slightly above the $852 billion figure implied by its most recent funding round. In other words, the market that trades these illiquid, insiders-only shares is now saying Anthropic is the more coveted bet.

If you zoom out, the speed of this repricing is staggering. Three months ago, Anthropic was celebrating that $380 billion Series G. Since then, the tone around the company in investor circles has shifted from “promising rival to OpenAI” to “you will regret not owning this.” Venture firms, crossover funds, and family offices are blasting the inboxes of employees and early backers with offers, sometimes several times a day, trying to pry loose even small allocations. One secondary-market banker described it as “an epic run,” saying there are “almost no sellers” and that deals often get snatched up within a day of appearing.

Some of the anecdotes feel like peak bubble behavior. One Anthropic shareholder reportedly tested the market with a block of shares that would value the company at $1.15 trillion. A “very well known growth fund” floated a buy at a $1.05 trillion valuation, prompting the X post that helped this whole story explode: “Absolutely wild.” In at least one case, a would-be investor went as far as advertising their home for sale in exchange for Anthropic shares, at a sub-trillion but still eye-watering valuation north of $800 billion. It’s the kind of thing you’d expect to see at the top of a cycle, not in the careful footnotes of a secondary liquidity process.

So what’s fueling this frenzy, beyond hype? Underneath the headlines, Anthropic actually does have a monster growth story to sell. The company has leaned hard into enterprise products, particularly Claude Code, its AI coding assistant that plugs into developer workflows and cloud environments. Independent analyses and industry reports peg Claude Code at an estimated $2.5 billion ARR run-rate as of early 2026, after hitting the first $1 billion in just about six months. One breakdown estimated Anthropic’s total revenue has jumped roughly 14x between late 2024 and early 2026, with Claude Code already contributing around a fifth of that haul. For big institutional buyers, that’s exactly the kind of metric that makes a nosebleed valuation at least arguable.

There’s also the broader context: a corporate AI land grab where every Fortune 500 CIO is trying to pick a foundation model partner they can bet on for years. Anthropic has positioned Claude as the “aligned,” safety-first alternative to some of the more experimental offerings from rivals. Its research into constitutional AI and “agentic” systems has given it a reputation for discipline and reliability that plays well with risk-averse enterprises and regulators. That narrative, combined with aggressive go-to-market execution on coding and productivity workloads, has turned into a powerful story for investors who missed the earliest OpenAI rounds and are looking for the next compounding machine.

Meanwhile, OpenAI finds itself in the strange position of being both massive and, in this specific corner of the market, slightly out of favor. The company still carries a headline valuation in the $800 billion-plus range from its last capital raise, and by almost any normal standard, demand for its equity would be considered extraordinary. Yet secondary traders tell a different story: in 2026, bid interest in OpenAI shares has cooled, with some brokers saying demand is “tepid” and that blocks can be “almost impossible to unload” at previous prices. Some of that may be simple rotation—investors feel they already “missed” OpenAI and are chasing what looks like fresher upside at Anthropic.

An important nuance here: none of these companies are public, so all of these numbers are, in a sense, vibes. Secondary-market valuations are not the same as a price set in a fully liquid public market or a large primary funding round. They’re shaped by tiny slivers of float, aggressive buyers who are more motivated than sellers, and a heavy dose of FOMO. When there are “almost no sellers,” as one banker put it, even a small group of institutions can push the implied valuation sky high. This is part price discovery, part social signaling; as one investor admitted, sometimes it’s less about the prospective return and more about being able to tell LPs “we got into Anthropic.”

Still, markets are ultimately making a relative call: if you can buy one AI giant or the other at roughly similar levels, the flows right now say Anthropic looks more compelling. That preference is reinforced by a drumbeat of reports that Anthropic may pursue an IPO as soon as late 2026, with some coverage suggesting its official valuation could eventually catch up to these secondary prices if public-market appetite holds. Prediction and betting markets that track potential AI valuations have also started to factor in scenarios where Anthropic, not OpenAI, is the first big foundation model player to crack a multi-trillion-dollar public market cap over the next cycle.

For the people inside Anthropic, this moment is both vindication and a pressure cooker. Early investors like Bradley Horowitz say they’re getting “daily offers from the ridiculous to the sublime” but have little interest in cashing out, insisting they’re “playing a long game.” Employees sitting on paper gains now worth tens of millions of dollars have to decide whether to sell a slice in a frothy market or hold on in the hope that an IPO or later round will validate those trillion-dollar marks. And leadership has to navigate all of this without appearing intoxicated by the numbers, especially as regulators and policymakers scrutinize AI companies for both competition and safety risk.

If you’re watching from the outside, it’s tempting to call this the top—classic late-stage startup behavior, exotic secondary deals, houses being traded for shares in a company that doesn’t yet trade on any stock exchange. But it’s also possible we’re still early in a genuine platform shift where a handful of AI infrastructure players do end up worth multiple trillions over time, much like cloud, mobile, and search platforms before them. The truth is, nobody really knows whether a trillion-dollar private price for Anthropic is genius or madness. What we can say is that, at least for now, one corner of the market has made its choice: in the battle of the AI unicorns, Anthropic, not OpenAI, is the name investors are chasing hardest.