The OpenAI valuation reached $852 billion in April 2026. That number is larger than the GDP of the Netherlands. It is larger than the combined GDP of Portugal, Greece, and Hungary. It makes OpenAI the most valuable private company in history, surpassing every startup that came before it by a margin that would have seemed fictional five years ago.
OpenAI has never made a profit. Not in 2023. Not in 2024. Not in 2025. And according to its own internal projections, it will not make a profit in 2026 either, a year in which it expects to lose approximately $14 billion. Over the period from 2023 through 2028, the company projects cumulative losses of $44 billion before turning its first profit in 2029. A separate internal projection puts cumulative losses through 2029 at $115 billion. The range between those figures reflects genuine uncertainty about the path ahead.
This is the central paradox of the most consequential technology company of the current decade: a business that burns money at a pace without precedent in corporate history, valued as though it has already won a prize nobody has officially awarded yet.
The Numbers Behind the Paradox
The financial profile of OpenAI in 2026 is unlike anything in the history of technology investment.
On the revenue side, the company has grown at a rate that challenges description. Monthly revenue of $300 million in August 2024, already representing a 1,700% increase since early 2023, had grown to $2 billion per month by March 2026. That trajectory, if sustained, points toward annual revenue of $24 billion by the end of 2026 — a figure that would place OpenAI among the largest software companies in the world by top-line revenue alone.
The problem is what sits beneath that top line. Compute infrastructure costs the company between $7 and $9 billion per year, primarily in GPU clusters accessed through Microsoft Azure. Salaries and equity compensation add more than $2.5 billion annually. Research and development spending runs at a level that reflects the company’s position at the absolute frontier of AI capability development. The burn rate is projected to remain at 57% of revenue through 2027, meaning that for every dollar OpenAI earns, it spends roughly $1.57.
In the first funding round, which closed in February 2026, OpenAI raised $110 billion at an $840 billion post-money valuation, backed by Amazon, Nvidia, and SoftBank. It was the largest private funding round in history, exceeding the previous record by a factor that would have seemed implausible a decade ago. The round provides approximately three to four years of operational runway at current burn rates, assuming the conditional tranches and compute credits within the round are fully realised.
The full financial context for OpenAI’s funding history and valuation trajectory is documented through a detailed analysis of OpenAI’s financial documents, including internal projections obtained by technology journalists covering the company’s capital raises.
Why Investors Are Betting Hundreds of Billions on a Money-Losing Company
The conventional framework for valuing a company — revenue, profit, cash flow, multiples — does not straightforwardly apply to OpenAI. Understanding why requires understanding what investors believe they are actually buying.
The thesis behind the OpenAI valuation is not that the company will eventually become a profitable software business in the traditional sense. It is something considerably more ambitious and considerably harder to price: the belief that general-purpose AI infrastructure will become the foundational layer of the global economy, and that whoever controls the leading position in that infrastructure will be able to extract value from it at a scale that makes current losses trivial in retrospect.
The historical precedent investors point to is Amazon Web Services. AWS was a loss-making internal experiment for years before it became the most profitable division of one of the world’s most valuable companies, generating margins that subsidised Amazon’s entire retail operation. The argument for OpenAI is structurally similar: the infrastructure being built today is expensive and unprofitable now, but its value compounds as the ecosystem built on top of it grows.
SoftBank, Amazon, and Nvidia collectively put more than $150 billion into OpenAI across two years on exactly that thesis. Nvidia’s commitment of up to $100 billion is particularly notable — and particularly complex. As OpenAI’s CFO, Sarah Friar acknowledged, that investment “will go back to Nvidia” in GPU purchases. Nvidia is also a significant investor in CoreWeave, which supplies cloud computing capacity to OpenAI and has itself spent billions buying Nvidia chips. The circular structure of these arrangements is not incidental; it reflects a broader pattern in AI investment where the companies selling the infrastructure are also funding the companies consuming it.
The competitive dynamics add urgency to the investment thesis. ChatGPT’s web traffic share fell from 86.7% in January 2025 to 64.5% in January 2026, a 22-percentage-point decline in 12 months, while Google Gemini captured much of the loss, growing from 5.7% to 21.5%. An Apple partnership integrating Gemini into Apple Intelligence accelerates that shift. Investors betting on OpenAI are betting on a company whose market position, while dominant today, is already under meaningful pressure from competitors with significantly deeper resources.
The Structural Problem With the OpenAI Valuation
Beneath the surface of the extraordinary OpenAI valuation lies a structural challenge that its financial projections acknowledge but cannot fully resolve.
The path to profitability as currently modelled depends on revenue growing to approximately $100 billion annually by 2029 — a figure that would place OpenAI alongside Nvidia at the peak of its semiconductor boom. Achieving that in four years would require compounding the current revenue base at rates that have few historical precedents for a company at this scale. The projection is not impossible, but it requires a sequence of assumptions about market growth, competitive retention, pricing power, and cost reduction, each of which has to hold simultaneously.
The cost structure presents its own challenge. Compute costs are not fixed. As OpenAI builds more powerful models, training runs become more expensive. The GPT-4 training run was estimated to cost more than $100 million. Future models at the frontier of capability will cost more. Research costs scale with ambition, and OpenAI’s stated ambition is to build artificial general intelligence, a goal whose scope makes cost forecasting genuinely difficult.
There is also the question of structural transformation. OpenAI converted from a capped-profit structure under nonprofit governance to a for-profit entity in 2025, a transition that unlocked the scale of external investment the company needed but also changed the incentive structure in ways that its original mission-driven culture may not have fully absorbed. An IPO is expected in 2027 or 2028, at which point the OpenAI valuation will face scrutiny from public markets that operate on different timescales and different tolerances for losses than the private investors who have funded the company to date.
What OpenAI Actually Earns and Where
The revenue picture is real, even if the profits are not. By February 2026, ChatGPT had reached 900 million weekly active users — a figure that represents growth from 100 million users in November 2023, adding 800 million users in roughly 27 months. Almost no consumer platform in history has scaled at that pace.
The revenue breakdown reflects two distinct business models running in parallel. Consumer subscriptions, primarily ChatGPT Plus at $20 per month, provide a large but relatively modest revenue base given the free tier’s scale. Enterprise licensing and API access represent the higher-margin segment, with businesses integrating OpenAI’s models into products, workflows, and services at rates that reflect the genuine productivity value the technology delivers.
The API business is where the long-term economic logic becomes clearest. When a company builds a customer service application, a coding assistant, or a document processing tool on OpenAI’s models, it creates a dependency that is costly to reverse. Switching costs are real in enterprise software. The more deeply OpenAI’s technology is embedded in the workflows of large organisations, the more defensible its revenue base becomes. This is the flywheel investors are trying to fund into existence.
The competitive erosion visible in ChatGPT’s traffic share data suggests the consumer market may be less defensible than the enterprise segment. Google, Anthropic, Meta, and others are all competing aggressively for consumer AI usage with products that are, for most everyday tasks, broadly comparable to OpenAI’s offerings. The brand advantage that ChatGPT accumulated by being first is real but not necessarily permanent.
This pattern of enormous capital deployment in pursuit of winner-take-most infrastructure positions connects to broader dynamics explored in how private capital is reshaping the industries people depend on and the structural risks that can follow when financial logic outpaces operational reality.
The Comparison That Puts the OpenAI Valuation in Context
The $852 billion OpenAI valuation is best understood not in isolation but in comparison to the companies it aspires to resemble.
Microsoft, which has invested approximately $13 billion in OpenAI and integrated its technology across its product suite, is valued at roughly $3 trillion. Google’s parent company, Alphabet, OpenAI’s most direct competitive threat in AI infrastructure, is valued at approximately $2 trillion. Nvidia, whose chips make everything OpenAI does possible, crossed $3 trillion in market capitalisation in 2024.
OpenAI at $852 billion is being valued at roughly a quarter of Microsoft, despite generating a fraction of Microsoft’s revenue and none of its profit. The implied multiple on current revenue is astronomical by any conventional measure. On projected 2029 revenue of $100 billion, if that projection holds, the multiple is more plausible but still aggressive relative to the uncertainty involved.
The honest answer to what the OpenAI valuation reflects is that it is not primarily a valuation of the business as it exists today. It is a valuation of a possible future in which AI infrastructure becomes as essential as cloud computing, and in which OpenAI occupies a position in that infrastructure analogous to where Amazon Web Services sits in cloud today.
That future may arrive. The bet is not obviously wrong. But the distance between the valuation and the current financial reality is large enough that it demands exactly the kind of scrutiny that public markets will eventually apply, and that the next few years of the company’s trajectory will begin to resolve.
The detailed financial projections behind OpenAI’s expected path to profitability are documented through analysis of internal documents that have been reviewed by financial journalists and published with appropriate verification.
Looking Ahead
The next three years will be among the most consequential in OpenAI’s history, and possibly in the history of technology investment.
If the company achieves something close to its projected revenue trajectory, the losses become manageable, and the valuation begins to look prescient rather than reckless. An IPO in 2027 or 2028, at revenue levels that justify the current private market price, would represent one of the most remarkable financial stories in the history of startups.
If the competitive pressure from Google, Anthropic, and others intensifies further, if the cost of frontier model training proves harder to reduce than projected, or if enterprise adoption grows more slowly than the revenue models assume, the gap between the OpenAI valuation and operational reality will become difficult to bridge.
The company is spending at a rate the Manhattan Project would have found notable, and projecting returns that would rank it among the most profitable businesses in human history within a decade. One of those things is observable. The other is a bet on a future that has not yet arrived.
For investors who have committed to that bet at prices that fully embed it, the next few years will determine whether the OpenAI valuation represents the most important capital allocation decision of the AI era, or its most instructive cautionary tale.
That question connects to a broader reckoning explored in how the AI investment boom is reshaping the global economy in ways that are only beginning to be understood, and why the race to control AI infrastructure is producing financial structures that challenge conventional economic logic.
