OpenAI has confidentially filed for an IPO that could rank among the largest in U.S. history, but the headline valuation masks a much more complex bet. Understanding what it is actually wagering on is one of the most instructive exercises in AI literacy right now.
Key takeaways
OpenAI has confidentially filed for an IPO that could rank among the largest in U.S. history, but key financial details remain private until a public S-1 is released.
The company completed a major restructuring in October 2025, becoming a Public Benefit Corporation controlled by its nonprofit foundation, a governance structure with no real precedent in large-cap public markets.
Revenue has grown dramatically, but OpenAI remains deeply unprofitable and does not expect to turn cash flow positive until around 2029 or 2030, making the IPO a long-duration bet on future dominance.
The Stargate project, a $500 billion AI infrastructure joint venture, is the physical foundation of OpenAI's growth narrative, linking compute capacity directly to revenue potential.
The shift from consumer chatbot to agentic enterprise platform is the core commercial bet: enterprise revenue now exceeds 40% of total revenue and is projected to reach parity with consumer by the end of 2026.
Sometime in the not-too-distant future, ordinary investors will be able to buy a small piece of the company that made ChatGPT a household name. That moment has been building for years, and it is now close enough to study seriously. OpenAI has confidentially filed for an initial public offering, setting itself up for what may be the most highly anticipated market debut in recent history. But the interesting question is not whether the IPO will happen. The interesting question is what, exactly, the whole thing is a bet on.
From Nonprofit to Public Benefit Corporation: The Corporate Archaeology Matters
To understand the IPO, you have to start with the corporate structure, because it is genuinely unusual. OpenAI was founded in 2015 as a nonprofit. For years it operated under a capped-profit model that limited how much investors could earn. That structure made sense for a research lab but was increasingly awkward for a company racing to raise tens of billions of dollars.
On October 28, 2025, OpenAI completed a restructuring that changed everything. The company's for-profit LLC transitioned to a Public Benefit Corporation (PBC), a purpose-driven company structure that has to consider the interests of both shareholders and the mission. The nonprofit branch, now called the OpenAI Foundation, did not disappear. Instead, it retained control and received a large equity stake in the new commercial entity, valued at roughly $130 billion at the time of the recapitalization. Microsoft, OpenAI's longtime backer, holds approximately 27% of the new structure, while current and former employees and other investors hold the remaining shares.
This restructuring lifted fundraising restrictions and gave OpenAI a clearer route to going public. Without it, the old capped-profit structure was difficult to map onto a conventional IPO. The PBC form is now the standard structure for other frontier AI labs as well, which means the IPO, if and when it completes, will list a genuinely novel kind of public company: one where a nonprofit controls the board of a for-profit entity with a near-trillion-dollar valuation.
The Revenue Story Is Real, and the Loss Story Is Also Real
Numbers are where the story gets both exciting and humbling. OpenAI's revenue trajectory has been extraordinary by almost any historical comparison. The company went from roughly $3.7 billion in annual revenue in 2024 to surpassing $20 billion in annualized revenue by the end of 2025, according to a blog post from CFO Sarah Friar. Audited 2025 financials, which surfaced ahead of the IPO preparations, showed full-year revenue of approximately $13 billion, representing roughly 253% year-over-year growth from 2024.
That growth is real. But so are the costs. Those same audited financials show that OpenAI spent approximately $34 billion in 2025 while generating $13 billion in revenue, resulting in a very large operating loss. The company does not expect to reach profitability until around 2030. Internal projections suggest losses of roughly $14 billion in 2026 alone. HSBC analysts have estimated that OpenAI may need well over $200 billion in additional funding by 2030, even accounting for projected revenue growth.
This is not a bug in the OpenAI story. It is the central feature of it. The IPO is not a bet that OpenAI is profitable today. It is a bet that the massive spend today buys a durable, dominant position tomorrow.
Stargate: The Infrastructure Wager Underneath Everything
The single biggest concrete bet embedded in the OpenAI IPO story is a project called Stargate. Announced on January 21, 2025, Stargate is a new company that intends to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States. The initial equity funders are SoftBank, OpenAI, Oracle, and MGX, with SoftBank holding financial responsibility and OpenAI holding operational responsibility.
The scale is almost incomprehensible. The flagship Abilene, Texas campus alone is expected to ultimately scale past a gigawatt of capacity, which is enough electricity to power about 750,000 U.S. homes. By September 2025, OpenAI announced five new data center sites and said the program had already secured nearly 7 gigawatts of planned capacity and over $400 billion in investment commitments, putting it on track to secure the full commitment ahead of the original schedule.
For any IPO prospectus, Stargate is a central question mark. Investors will need to understand whether Stargate is a supplier relationship, a joint venture, a financing vehicle, or some combination of all three. Who owns the assets being built? Who bears the construction risk? How does compute contracted through Stargate interact with OpenAI's existing Microsoft Azure relationship? These are questions that the public S-1 filing, when it arrives, will need to answer. Right now, OpenAI has confidentially filed its registration statement, meaning audited financials and full risk disclosures remain private. The answers will matter enormously to any rational assessment of what the company is worth.
What is clear is the thesis: OpenAI CFO Sarah Friar has pointed out that revenue growth closely tracked the expansion of computing capacity, which grew from 0.6 gigawatts in 2024 to 1.9 gigawatts in 2025. The implicit argument is that more compute means more revenue. If that relationship holds, Stargate is not just an infrastructure project. It is the engine of the entire growth story.
The Agentic AI Bet: From Chatbot to Operating System
The subscription chatbot model that drove ChatGPT's initial growth is not the primary bet in the IPO. The primary bet is on agentic AI: software that does not just answer questions but takes autonomous actions across business systems on behalf of users.
OpenAI has been transparent about this ambition. Enterprise revenue now makes up more than 40% of total revenue, and the company has said it is on track to reach parity with consumer revenue by the end of 2026. OpenAI's Codex coding agent crossed 3 million weekly active users, a figure that was described as near zero at the start of the same quarter. Paying business users reached 9 million in February 2026, up from 5 million in August 2025.
OpenAI introduced a dedicated enterprise platform called Frontier, designed to help organizations build, deploy, and manage AI agents that can operate across business systems. The company has formalized integration partnerships with McKinsey, Boston Consulting Group, Accenture, and Capgemini to help enterprises deploy agents at scale. OpenAI projects reaching $85 billion in revenue by 2030, a number that, in the company's own framing, only makes sense if agents become the default way businesses interact with AI, not just a feature layered on top of a chat interface.
This matters for the IPO because an agentic revenue model, where AI systems complete tasks and workflows rather than simply answer prompts, implies much higher revenue per user than a subscription chatbot. It also implies much deeper lock-in. An enterprise that has woven AI agents into its finance, customer service, and supply chain systems is significantly harder to displace than one that simply pays a monthly fee for a chatbot.
The Competition Problem Public Markets Will Have to Price
Buyers of any future OpenAI shares will be purchasing stock in a company that faces more credible competition than it did two years ago. Google's AI products have grown their web traffic share significantly, while ChatGPT's share of AI-assistant web traffic has declined over the same period. Anthropic, which recently raised funding at a valuation that briefly surpassed OpenAI's, is targeting its own IPO. Meta is investing heavily in open-source frontier models, and open-source alternatives put persistent downward pressure on API pricing across the industry.
In enterprise AI specifically, the competitive picture has shifted. According to Menlo Ventures data from late 2025, Anthropic holds approximately 40% of enterprise LLM API spend, while OpenAI's share had dropped to 27%, down from roughly 50% in 2023. That is a meaningful reversal for a company whose IPO valuation depends on compounding enterprise dominance.
The structural risk is the capital intensity of staying at the model frontier. Maintaining a lead requires continuous spending on compute, data, and talent, with no guarantee that any single model generation sustains pricing power. Open-source alternatives and well-funded rivals compress margins from below, while heavy reliance on cloud and chip suppliers limits cost control.
Valuation: What Does a Near-Trillion-Dollar Price Tag Require?
The valuation figures attached to OpenAI are large enough to cause genuine vertigo. Private market transactions have valued the company at approximately $852 billion following a March 2026 funding round. Some reports have suggested the IPO itself could target a valuation of $1 trillion or more. For context, Meta was valued at $104 billion when it went public in 2012, and Uber at $82 billion in 2019.
At an $830 billion valuation, OpenAI's price-to-sales ratio would be roughly 65 times 2025 revenue, far higher than most technology companies at any stage of their public life. That multiple is only defensible if investors believe the revenue growth continues at a dramatic pace for many years, that the path to profitability after 2029 or 2030 is genuine, and that OpenAI maintains or expands its competitive position against well-funded rivals.
The IPO is also reportedly targeting a raise of approximately $60 billion, which would more than double Saudi Aramco's 2019 listing of $25.6 billion, the largest IPO on record at that time. A $60 billion raise would also be larger than the total raised in U.S. IPOs in all but a handful of years since 1980. These are not just big numbers. They are a structural test of whether public equity markets are willing to fund frontier AI development at the scale that OpenAI believes is necessary.
The Governance Puzzle That No Tech IPO Has Tried Before
Beyond revenue, costs, and competition, the OpenAI IPO presents a governance question that has no clear precedent in the history of major technology listings. A nonprofit foundation, holding approximately 26% of the equity, retains the authority to appoint the board of the public benefit corporation. Microsoft holds roughly 27% but has zero governance control. Employees collectively hold significant equity but limited voting power on major decisions.
This means ordinary public shareholders who buy shares after the IPO will own an economic stake in a company whose strategic direction is ultimately guided by a nonprofit foundation whose mandate is to ensure artificial general intelligence benefits all of humanity. Public benefit corporations have a legal obligation to consider shareholder interests alongside their public benefit mission, but they are not legally required to prioritize public benefit over profit. The balance between those two commitments is untested in a company of this scale.
For investors trying to reason about this, the practical question is whether the foundation's mission-driven oversight makes OpenAI more or less likely to take short-term profits over long-term research bets. Historically, the governance structure has allowed OpenAI to sustain losses at a scale that purely profit-driven organizations would not tolerate. Whether that turns out to be an asset or a liability for public shareholders depends on how well the long-term bets pay off.
The Agent5 Angle: Making Probabilistic Predictions About What Happens Next
Getting smart about AI means developing the habit of treating future outcomes as probabilities rather than certainties. The OpenAI IPO is a useful case study in that discipline.
Consider the range of outcomes. In one scenario, the compute-to-revenue relationship holds, Stargate delivers the infrastructure advantage OpenAI expects, and agentic AI becomes the backbone of enterprise operations by the early 2030s. In that world, a near-trillion-dollar valuation looks prescient in retrospect. In another scenario, open-source models narrow the capability gap, enterprise customers resist deep lock-in, and Stargate's extraordinary capital requirements weigh on profitability for longer than projected. In that world, public shareholders who bought in at a high multiple face years of underperformance even if the company remains dominant.
The point is not to pick one of those futures with certainty. The point is to identify which variables matter most, then watch them. Key signals to track: Does OpenAI's share of enterprise AI spending recover or continue to erode? Does the compute-to-revenue relationship hold as Stargate brings more capacity online? Does the nonprofit governance structure produce any visible friction with commercial priorities once public shareholders are in the room? And does the broader AI market grow fast enough to lift all boats, or does it compress into winner-take-most dynamics that reward only one or two players?
An IPO filing is, among other things, a moment of forced transparency. When the public S-1 eventually arrives, it will be one of the most information-rich documents produced about the AI industry to date. Reading it carefully, not just the headline valuation, is among the most educational things a curious observer of AI's trajectory can do.
OpenAI has filed a confidential S-1 registration statement with the SEC, which begins the formal review process but does not yet make financial details public. The company has not set a final IPO date, price, or share count. Under SEC rules, the full registration statement must be made public at least 15 days before any IPO roadshow begins.
A Public Benefit Corporation is a for-profit structure that legally requires the company to balance shareholder financial interests with a defined public mission. OpenAI chose this structure because it preserves a connection to its original mission of ensuring AI benefits humanity, satisfies regulators in California and Delaware who were involved in approving the restructuring, and mirrors the structure used by other frontier AI labs. Crucially, OpenAI's nonprofit foundation retains control and a large equity stake in the PBC.
Stargate is a joint venture announced in January 2025, intending to invest $500 billion over four years building AI data center infrastructure for OpenAI across the United States and internationally. OpenAI holds operational responsibility while SoftBank holds financial responsibility. It matters for the IPO because OpenAI's CFO has publicly linked revenue growth directly to expansions in computing capacity, making Stargate effectively the engine of the company's growth narrative.
No. The company does not expect to reach profitability until around 2029 or 2030. Audited 2025 financials showed the company spent roughly $34 billion while generating approximately $13 billion in revenue. Internal projections suggest losses of roughly $14 billion in 2026. The bet embedded in the IPO is that sustained high spending today buys a dominant competitive position that becomes profitable at scale later.
The most direct competitors are Anthropic, which recently raised funding at a valuation that briefly surpassed OpenAI's and is also planning its own IPO, and Google, whose AI products have grown their share of AI-assistant web traffic. Meta's open-source model investments also create indirect competitive pressure by pushing down API pricing across the industry. In the enterprise market specifically, Anthropic's share of LLM API spending has grown significantly while OpenAI's has declined from its earlier peak.
The public S-1 will be the most detailed financial and strategic document OpenAI has ever published. Key things to examine include the gross margin on different revenue streams (subscriptions versus API versus enterprise), the precise terms of the Stargate and Microsoft relationships and what financial obligations they create, the actual path to profitability including specific assumptions, and the governance rights of public shareholders relative to the nonprofit foundation. The risk factors section will also describe regulatory and competitive threats in unusually candid terms.
