Anthropic was supposed to file in October. Instead, the company pushed its planned listing to November, and the reason it gave was almost disarmingly practical: it wants one more quarter of numbers on the table before it asks the public market for what could be the largest technology offering in history. That is not the move of a company nervous about its growth. It is the move of a company that thinks its growth is the whole pitch, and wants investors staring at as much of it as possible before they sign the check.
The number attached to that pitch is $2 trillion. According to reporting from The Wall Street Journal and confirmed across multiple outlets, Anthropic and its bankers are targeting a valuation around that figure, alongside a primary raise of up to $100 billion, a sum that would eclipse SpaceX’s record private placement and every tech IPO in the last 45 years combined. Whether the number survives contact with public markets in November is genuinely an open question. The case for it, and against it, comes down to a single chart.
The short version
- Anthropic delayed its planned IPO from October to November so it can present Q3 financials before the roadshow, according to The Wall Street Journal
- Bankers, including Morgan Stanley, Goldman Sachs and JPMorgan, are pitching the offering at a $2 trillion valuation with a raise of up to $100 billion
- Anthropic’s annualized revenue run rate hit $65 billion by the end of July 2026, up from roughly $47 billion in May and about $9 billion at the end of 2025, a sevenfold jump in seven months
- The $2 trillion figure is reportedly underwritten against projected 2028 revenue of $190 to $200 billion, a number that has not happened yet
- Skeptics, including analysts at ION Analytics, estimate the valuation would require $59 to $79 billion in annual profit to justify itself at multiples comparable to other Nasdaq 100 companies
- The timing lands awkwardly next to CEO Dario Amodei’s public calls for the industry to slow down, a tension Anthropic has not fully addressed
The chart doing the arguing
Start with what is not in dispute. Anthropic’s revenue has grown at a pace that has no real precedent in enterprise software. The company ended 2025 with annual revenue of roughly $9 billion, then told investors its annualized run rate had climbed to about $47 billion by May 2026, then to $65 billion by the end of July. A preliminary second-quarter revenue figure of $11.5 billion represented a fourteenfold jump from the same quarter a year earlier. Investors briefed on the numbers reportedly expect the run rate to clear $110 billion by the end of 2026.
That trajectory is the entire argument for a $2 trillion valuation, and it is a genuinely remarkable one on its own terms. What it is not, yet, is $190 to $200 billion in annual revenue, which is reportedly the figure bankers are using to underwrite the deal, a 2028 projection rather than a 2026 result. A valuation built on where a company is expected to be in two years, rather than where it is now, is not unusual for a fast-growing tech IPO. It is unusual for the gap between “expected” and “current” to be this large, roughly three times the run rate Anthropic is reporting today.
Where the number actually comes from
The $2 trillion figure would value Anthropic at more than double its last disclosed private funding round, and at roughly 30 times its July run rate rather than its trailing revenue, a multiple that only makes sense if 2028 arrives on schedule. Morgan Stanley, Goldman Sachs and JPMorgan are reportedly leading the underwriting, and the raise itself, up to $100 billion, would be large enough on its own to fund years of the compute spending that AI labs have made into a permanent cost center. Anthropic’s cloud footprint has expanded accordingly; the company’s infrastructure commitments this year alone run into the tens of billions across its cloud partners, spending that only makes sense against a revenue base expected to grow several times over.
| Milestone | Figure |
|---|---|
| Annual revenue, full year 2025 | ~$9 billion |
| Annualized run rate, May 2026 | ~$47 billion |
| Annualized run rate, July 2026 | ~$65 billion |
| Investor estimate, run rate by year end 2026 | $110 billion+ |
| Projected annual revenue used to underwrite IPO, 2028 | $190 to $200 billion |
| Target IPO valuation | ~$2 trillion |
| Target primary raise | Up to $100 billion |
For scale, SpaceX, OpenAI and Anthropic’s combined private valuations already exceed the first-day market capitalization of every U.S. tech IPO since 1980. A $2 trillion Anthropic listing would put it in the same tier as Apple, Microsoft and Nvidia on day one, before a single quarter of public reporting, for a company that generated single-digit billions in revenue less than two years ago.
What the skeptics are actually arguing
The pushback is not that Anthropic’s growth is fake. Nobody disputes the run-rate numbers themselves. The argument is about durability and price. Analysts at ION Analytics have estimated that justifying $2 trillion at multiples typical of Nasdaq 100 companies would require Anthropic to generate somewhere between $59 billion and $79 billion in annual profit, not revenue, profit, a bar the company has not disclosed hitting and that would require margins well above what AI labs have historically run given the cost of the underlying compute.
Gil Luria, an analyst at D.A. Davidson, was blunter about the underlying skepticism, describing the IPO push as something that “feels more and more like a ladder pull,” the implication being that Anthropic and its backers want to cash out at a peak valuation before growth inevitably normalizes. Forbes ran a piece questioning whether a structure this large, with insiders and existing investors holding outsized influence over allocation, put ordinary public investors at a structural disadvantage from day one. On the other side, Menlo Ventures’ Matt Murphy, an existing Anthropic investor with an obvious stake in the outcome, called the growth rate “off the charts” and said he sees no near-term reason for it to slow.
There is also a pricing question that cuts closer to Anthropic’s actual product. Claude’s flagship pricing runs meaningfully above OpenAI’s comparable tier, roughly two and a half times as expensive by some estimates, even as Anthropic has been cutting costs elsewhere, most notably a 75 percent reduction in cached-token pricing on its newest model in September. That cut targets the repeated-context workloads that make agents expensive to run, the same category of spending Google has been chasing with Gemini’s own aggressive Flash-tier pricing. Anthropic is competing on price at the margins while its flagship model stays expensive, a strategy that works only if enterprise customers keep paying a premium for Claude specifically, rather than routing routine work to whichever model is cheapest that month.
The awkward part nobody has fully explained
The timing sits uneasily next to Anthropic’s own public position on AI risk. CEO Dario Amodei spent much of 2026 publicly urging the industry to slow down frontier development to avoid what he has described as catastrophic risk, and by September, OpenAI, Anthropic and Google DeepMind had reportedly been coordinating privately on safety measures for weeks, according to OpenAI policy chief Chris Lehane. A company whose leadership is publicly arguing for restraint is simultaneously asking public markets to fund a $100 billion capital raise built on the premise that its own growth needs to keep accelerating for years. Those two positions are not technically contradictory, a company can urge industry-wide caution while still competing hard within whatever guardrails exist, but Anthropic has not spent much effort reconciling them for the investors it is about to ask for money.
Signals to watch
- Q3 numbers, once filed. The entire reason for the delay was to show them before the roadshow, so whatever they say will move the final valuation more than anything reported so far
- Whether the $190 to $200 billion 2028 projection gets revised. Bankers routinely adjust underwriting assumptions as a roadshow progresses and investor appetite becomes clearer
- How OpenAI’s own IPO plans respond. OpenAI has been notably quieter about a near-term listing while Anthropic pushes forward, a divergence worth watching given how closely the two companies are usually compared
- Whether margin, not revenue, becomes the story public investors actually focus on. Revenue growth got Anthropic here. Profit is what a public market will spend the next several years asking about
None of this means $2 trillion is an irrational number, or that it is the right one. Anthropic’s revenue chart is real, and growth at this pace, sustained for even a few more quarters, would justify a great deal of investor enthusiasm on its own. But an IPO is where a growth story stops being told entirely on the company’s terms. Once Anthropic files, the run-rate chart stops being the only argument in the room, and the harder questions, about margin, about pricing power, about whether 2028’s number actually materializes, get asked by people with money on the line rather than reporters relaying what bankers say off the record. November will be the first real test of whether the chart is enough.
