THE LAST TIME two technology companies of this magnitude raced each other to the public markets, the year was 1999 and the companies were selling pet food and groceries online. The stakes are considerably higher now. Anthropic, the maker of Claude, has begun discussing an initial public offering as early as the fourth quarter of 2025, with bankers expecting the company to raise upward of $60 billion — a figure that would make it the second-largest IPO in history, trailing only SpaceX's anticipated $75 billion listing. Its archrival OpenAI, last valued at $730 billion, is maneuvering to beat Anthropic to the exchange, with Sam Altman reportedly telling associates he would prefer ChatGPT's parent to list first.
The financials underpinning both companies have shifted dramatically in recent months. Anthropic's annualized revenue more than doubled during the first two months of 2026, reaching $19 billion on the strength of its automated coding tools and enterprise developer platform. OpenAI, meanwhile, topped $25 billion in annualized sales last month, driven primarily by ChatGPT subscriptions and its expanding consumer footprint. Anthropic was last valued at $350 billion in a financing round completed just weeks ago; it has retained Wilson Sonsini as legal counsel for the offering, a concrete step that suggests the IPO discussion has moved well beyond the hypothetical. OpenAI, for its part, has held informal conversations with Goldman Sachs and Morgan Stanley, though with $120 billion in fresh capital nearly closed, it faces no immediate financial pressure to list.
Public offering, private ambitions
But the race to go public is not merely a contest of vanity metrics or founder egos — it is a referendum on which business model the public markets believe will define the AI era. Bankers and IPO advisers largely expect Anthropic to list first, and the reasoning is instructive: institutional investors, they reckon, favor Anthropic's enterprise-and-developer focus over OpenAI's consumer-heavy revenue mix. A company selling picks and shovels to the industry tends to inspire more confidence on a roadshow than one whose largest product is a $20-per-month chatbot subscription, however ubiquitous that chatbot may be. Anthropic has also projected a shorter path to positive free cash flow — 2028, versus OpenAI's 2030 — and a considerably smaller cumulative cash burn to get there: roughly $22 billion compared to OpenAI's projected $200 billion-plus.
Yet neither company's path to profitability is anything close to assured, and the revenue accounting alone will give securities regulators plenty to chew on. Both Anthropic and OpenAI generate substantial revenue through cloud providers — Amazon Web Services and Google Cloud for Anthropic; Microsoft Azure for OpenAI — that resell their models to end customers. How each company books that revenue, and how much of it represents durable enterprise demand versus cloud-credit consumption that could evaporate when contracts expire, will almost certainly become a focal point of SEC scrutiny. The companies have also projected revenue growth rates that are, by any historical standard, unprecedented; whether public market investors will underwrite those projections at current valuations is an open question that neither founder can answer in advance.
The macro backdrop adds another layer of uncertainty. Rising oil prices tied to the Iran conflict have stoked inflation fears and dampened consumer confidence — precisely the conditions that tend to freeze IPO windows shut. Both companies are also pursuing novel strategies to lock in enterprise customers before they go public: Anthropic is in talks to form a joint venture with private equity behemoths Blackstone, Hellman & Friedman, and Permira, while OpenAI has explored a similar arrangement with TPG, Advent International, Bain Capital, and Brookfield. These deals, if consummated, would embed AI tooling directly into the portfolio companies of the world's largest buyout firms — a distribution channel that could prove formidable, albeit one that blurs the line between organic adoption and contractually obligated usage.
The coming months will determine whether the public markets are ready to absorb two AI listings of this scale in quick succession, or whether one company's debut will siphon the oxygen from the other's. For all the talk of a $60 billion raise here and a $730 billion valuation there, neither Anthropic nor OpenAI has yet demonstrated that it can generate a single dollar of free cash flow. The AI boom has produced staggering growth, eye-watering capital deployment, and a pair of companies now valued above most of the Fortune 100. What it has not yet produced is a profit. Investors who buy in at these levels are making a wager not on what these companies are, but on what they might become — which, in fairness, is what IPOs have always been about. The difference is that the chips on the table have never been this large. ■
For more, join 75,000 subscribers getting tech's favorite brief here
