The Company That Created a Category
When ChatGPT launched in December 2022, it marked the first time the broader public encountered the genuine power of artificial intelligence. The company behind it not only introduced a product, it invented an entire category. For roughly two years afterward, that company sat comfortably at the top of the field, viewed by virtually every observer as the unambiguous leader in frontier models, in consumer chat applications, and increasingly at the enterprise level.
It is striking how quickly that perception has shifted. As recently as nine months ago, the conventional wisdom still positioned OpenAI as the dominant force in AI. Today, the picture is far more competitive, with Google's Gemini posting significant gains in the consumer market and Anthropic emerging as a particularly potent force in enterprise deployments.
A Funding Round Without Precedent
One element of OpenAI's story that has not received the attention it deserves is the sheer magnitude of its most recent capital raise. The company secured $122 billion in a single funding round — an amount that is not merely the largest funding round in history, public or private, but is roughly four times larger than the next biggest round on record. This is a quantum leap in private financing.
That capital matters, because the economics of frontier AI development are unforgiving. OpenAI has secured roughly $300 billion in compute commitments from Oracle and approximately $250 billion from Microsoft, alongside significant arrangements with Amazon and Coreweave. The relationship with Coreweave in particular is consequential because OpenAI represents a very large portion of that company's backlog. Yet OpenAI's compute footprint is far more diversified than any single supplier, reflecting both the scale of its ambitions and the prudence of not depending on a single source.
The Reality Behind the Headlines
A recent report suggesting OpenAI was failing to meet certain internal targets briefly rattled markets, but the unease did not last. The reporting was largely backward-looking, drawing on conversations that had taken place over the previous six months — precisely the period during which the competitive landscape had shifted underneath the company. Earlier expectations had been built on an assumption of unrivaled leadership, and as Gemini gained traction with consumers and Anthropic with enterprise customers, OpenAI's own trajectory had to be recalibrated.
Yet a sober look at the numbers tells a story of remarkable success rather than disappointment. The company moved from a $12 billion annualized run rate six months ago to roughly $24 billion annualized today. By any reasonable standard, this is exceptional performance. Even if early ambitions reached for the stars, the company has, at minimum, reached the moon — with every indication that it intends to continue climbing from there.
A Crowded Field at the Frontier
The current state of frontier AI is intensely competitive. Three firms lead the way: OpenAI, Google's Gemini, and Anthropic. In the consumer chat market, OpenAI retains the largest share, with Gemini a distant second and Anthropic a distant third. In the enterprise market, however, the order shifts: Anthropic now leads, OpenAI sits in second, and Gemini comes third.
And these three are far from the only contenders. Meta has clear ambitions to compete, particularly on the consumer side. Elon Musk's XAI is similarly focused on consumer applications. The largest hyperscalers — Microsoft and Amazon — also harbor their own model development plans at various levels. The result is a marketplace that is uncommonly crowded for one so young, with deep-pocketed entrants on every side. Despite this pressure, OpenAI remains very well positioned, anchored by the most successful consumer product in the category, a competitive enterprise offering, and a research lab whose pedigree continues to attract capital and talent.
The Coming Wave of Public Listings
Perhaps the most consequential question facing the industry is the timing and structure of upcoming public listings. Demand for these companies is, by any reasonable measure, off the charts. This is visible not only in the most recent venture rounds but in the appetite among retail investors, who have piled into whatever vehicles they can find to gain exposure. These are the fastest-growing companies of their scale in history.
Current expectations point to SpaceX going public first, which will also bring the XAI lab housed within it into the public sphere. Both OpenAI and Anthropic are likely to follow, with both probably aiming to list within the year. The motivation is straightforward: both companies are losing money at meaningful scale, and they need fresh capital to fund the compute resources required to train and serve increasingly capable models.
The mechanics of these listings will be unprecedented. Offerings of this magnitude have never been attempted before, and the logistical challenges are formidable — finding anchor investors, setting prices in the absence of meaningful comparables, and managing the rate of issuance at a scale the public markets have never absorbed. Compounding the difficulty, the accounting treatment for products of this kind has no clear precedent. The way these companies sell their offerings is genuinely novel, and standard definitions will need to be adapted or invented to fit.
What Investors Should Watch
For anyone trying to evaluate OpenAI as it transitions toward public ownership, the most important metric will be top-line growth: the share captured in consumer markets, the share captured in enterprise markets, the absolute size of each business, and the rate at which each is expanding.
A second, more elusive metric will matter just as much: a defensible measure of gross margin that allows apples-to-apples comparison with Anthropic. The central question is how much of every inference dollar ultimately flows to the bottom line. Reports have widely characterized these companies as profitable on inference, but estimates of gross margins span an enormous range — from roughly 30% to as high as 70%, depending on which definitions one accepts. That range is far too wide to support sound investment judgments.
This is, in fact, one of the genuine virtues of going public. A listing forces the adoption of consistent, GAAP-compliant reporting measures. Once the dust settles, investors will finally have a clear and uniform window into how profitable the leading AI labs really are at delivering their products — and that clarity will, in turn, shape how the entire industry is valued.
Conclusion
The story of frontier AI is no longer a story about a single dominant company. It is a story of three intensely competitive firms, each deeply capitalized, each pursuing the same scarce resources of compute and talent, and each preparing to test the public markets at a scale never before attempted. OpenAI may have created the category, but its leadership is now contested in ways that would have seemed implausible only a year ago. The coming year will determine not only which company emerges strongest, but how a financial system built for traditional businesses adapts to one of the most unusual industries the modern economy has ever produced.