IPO must get ahead of OpenAI; Anthropic aims to seize the “pricing power” in AI.

```

Anthropic has filed first, pushing the competition with OpenAI from models, revenue, and valuation to the hurdle of public market pricing.

On Monday, Anthropic announced that it had confidentially submitted its U.S. IPO application, getting ahead of rival OpenAI in entering the listing process. Anthropic said in a statement that submitting the prospectus "gives us the right to choose to list after the SEC completes its review," and emphasized that the "proposed initial public offering will depend on market conditions and other factors."

This is not yet a formal public offering. According to Reuters, Anthropic’s listing could happen as early as this autumn, but it has not disclosed offering size or terms. The meaning of confidential submission is that the company can prepare for an IPO without temporarily revealing sensitive financial details to competitors and the public.

This makes the IPO race among AI large model companies suddenly much more concrete. Previously, the market was more concerned with whose model was stronger and who had more users. Now, the question becomes: who will first accept public market scrutiny, and who will first price "cutting-edge AI companies."

This filing will test whether investor enthusiasm for AI can withstand scrutiny from the public market, and will determine which company first establishes a valuation template for the rapidly growing AI industry.

OpenAI has not followed up with its own filing for now. OpenAI CEO Sam Altman said he is "not focused on the timing of a potential IPO," and that the company "will go public when the time is right."

But the market wasn’t betting on this outcome. On prediction markets, most participants previously expected OpenAI to submit its IPO application ahead of Anthropic.

The IPO window is open, but money isn't unlimited

Anthropic is racing against time, backed by a very real market window.

Dealogic data shows that by May 26, global IPO financing had reached $87.5 billion, the highest level for the same period since 2021.

Currently, the IPO window has clearly opened. AI chip company Cerebras surged 68% on its first day of listing last month. According to FactSet data, among companies that went public with valuations over $10 billion in the past five years, only the digital design platform Figma saw a higher first-day gain at 250% last year.

But an open window does not mean unlimited capital.

It’s worth noting that SpaceX is also moving ahead with a major IPO, aiming to raise $75 billion at a valuation of $1.75 trillion, with trading possibly commencing in as soon as two weeks. If SpaceX, Anthropic, and OpenAI all list in close succession, the U.S. stock market will have to absorb several large-scale tech assets simultaneously.

IPOX Vice President Kat Liu told Reuters: "Filing not long after SpaceX allows Anthropic to take advantage of a still favorable window and leverage strong investor interest in AI and growth stocks."

She added: "Compared to SpaceX, Anthropic’s valuation demand doesn’t look as aggressive as it does on its own."

Issuer Network founder Patrick Healy said: "There’s only so much oxygen in the room." He added: "SpaceX will consume vast amounts of capital, so the second entrant will fare better than the third."

D.A. Davidson analyst Gil Luria said: "SpaceX, OpenAI, and Anthropic together will have extremely large capital demands, likely to disturb the capital markets, so listing earlier will be a huge advantage."

Filing first is grabbing the narrative, but also assumes risks first

The benefits of going public first are straightforward: setting the price first, raising funds first, providing liquidity for employees and early investors first.

But listing first also comes at a cost: first to publicly disclose finances, first to undergo scrutiny from institutional investors, first to expose the true cost structure of AI companies.

PitchBook senior analyst Harrison Rolfes said: "The conventional interpretation is that Anthropic has just secured a narrative advantage by filing first."

But he gave another view: "The unconventional interpretation is that OpenAI actually gets the better outcome: Anthropic voluntarily takes on all disclosure risks first, and OpenAI can now freely observe how institutional investors respond to audited frontier AI financial data before deciding its own pricing."

This statement hits the key to an AI company IPO. The public market will look not only at the "AI story," but also at revenue quality, computing costs, cloud service revenue sharing, cash burn, customer structure, and profit margins.

The Wall Street Journal, citing academic research, says IPOs often come in batches within an industry, and those companies that list later in the cycle typically do worse than early entrants. The report explains that companies with deeper moats and better quality tend to go public earlier, followed by a batch of followers.

But being first doesn’t guarantee profits.

In 2019, Lyft went public before Uber, but its post-IPO stock performance fell short of expectations, directly affecting Uber’s IPO two months later. Uber subsequently lowered its target valuation, but its stock still dropped after listing.

The same report cites that Facebook lost more than half its share price within three months of going public in 2012, as the market worried it could adapt to mobile advertising. Facebook later proved its business model, but other companies planning to go public, including Twitter, ultimately had to wait.

This means that Anthropic, by stepping up first, could secure the pricing power for the AI IPO, but might also become the first large model company whose books are "opened up" to the public market.

Why Anthropic: Revenue, valuation, and profit narratives are changing

Anthropic is now bold enough to move forward, thanks to financial changes over the past few months.

WallstreetCN mentions that Anthropic’s annualized revenue is close to $45 billion. OpenAI’s annualized revenue just broke $30 billion, now estimated at about $33 billion. By this measure, Anthropic’s revenue scale is at least 35% greater than OpenAI’s.

This shift happened rapidly. The report says by the end of 2025, Anthropic’s annualized revenue was only $9 billion, less than half of OpenAI’s. In the first five months of this year, Anthropic’s revenue grew about fivefold; at the same time, OpenAI’s revenue grew more than 50%.

The two companies' revenue structures also differ. OpenAI’s income mainly comes from ChatGPT subscriptions; Anthropic relies more on selling API access for AI programming and other white-collar work scenarios to enterprises.

For the public market, both types of revenue will be compared. Subscription revenue depends on user scale and retention, while enterprise API revenue is judged by customer stickiness, usage frequency, and unit economics.

In terms of valuation, Anthropic has already overtaken OpenAI.

Reuters reports that Anthropic completed a $6.5 billion funding round in late May, with a post-investment valuation of $96.5 billion, already surpassing OpenAI. OpenAI’s latest valuation in March this year was $85.2 billion.

Anthropic’s valuation growth has been rapid. In February this year, Anthropic’s fundraising of $3 billion valued it at $38 billion. By late May, its valuation had more than doubled. The latest investors include Blackstone, Brookfield, D1 Capital Partners, GIC, General Catalyst, and Insight Partners.

Anthropic’s swift rise early this year impacted software and IT stocks, as investors worried that more autonomous AI tools could alter traditional business models and accelerate industry disruption.

The profit margin is even more closely watched by the market.

According to The Information, Anthropic is expected to achieve about $559 million in operating profit in Q2, with about 5% operating profit margin.

OpenAI, meanwhile, remains deep in losses. The report says OpenAI’s Q1 operating loss rate stood at 122%, and this number is after stripping out significant items such as equity incentives. By this calculation, OpenAI’s quarterly operating loss was at least $7 billion.

The cost pressure comes mostly from computing power.

OpenAI estimated earlier this year that it would burn about $25 billion in cash for the year, with AI server rental costs as high as $32 billion. In addition, OpenAI must give Microsoft 20% of its total revenue under an agreement lasting until 2030. If this year’s revenue reaches the previously forecast $30 billion, revenue sharing with Microsoft would be about $6 billion.

Anthropic isn't free from cost pressure either. It also needs to share revenue with cloud partners. Its revenue figures include total sales via other cloud providers, part of which will ultimately be returned to those cloud partners.

Anthropic’s current profitability also faces risk. As revenue grows rapidly, the company will need to greatly expand server resources, which could plunge it back into losses.

This is what the public market will ask: How fast is revenue growing, are computing costs increasing faster or slower; how much of total revenue ultimately goes to partners; are enterprise customers genuinely retained, or just boosted by short-term AI fever.

This IPO race will ultimately become a public market stress test

From a timing perspective, Anthropic has already pushed the pace forward by a step.

From a financial narrative perspective, it also presents a mix more easily understood by the public market: higher annualized income, higher recent valuation, and at least short-term better operating profit performance.

But this does not mean the IPO outcome is settled. Confidential filing is not the same as a successful listing or finalized valuation. The true test will begin once the prospectus is made public.

The public market will compare Anthropic, OpenAI, and other AI companies side by side: revenue growth rate, profit margin, cash burn, computing expenditure, cloud sharing, customer structure, model capabilities, and commercialization path.

Who lists first and how the market reacts may affect both companies’ futures, and could shape the next stage of the AI boom: either reinforcing market confidence in AI’s transformative power or signaling a warning on excessive enthusiasm.

For investors, this competition is no longer just about "whose model is smarter." Now, it’s also about who can turn the AI story into financial statements the public market is willing to pay for first.

Risk Warning and DisclaimerThe market has risks, investment needs caution. This article does not constitute personal investment advice and does not take into account the special investment goals, financial situation or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions given in this article are suitable for their particular situation. If you invest accordingly, you bear your own responsibility. ```