OpenAI exhibits extreme revenue growth, projected to reach $13 billion by 2025 and $100 billion by 2028, but faces massive cash burn and is not expected to be profitable until 2030. This duality presents a significant investment enigma. Competitors like Anthropic show earlier profitability timelines. OpenAI’s significant capital outlays for AI research and infrastructure drive its sustained losses. Potential investors can gain exposure through Microsoft (its largest shareholder), Nvidia, Broadcom, Oracle, Reddit, or specialized funds and ETFs. The company's IPO horizon in 2026 involves complex ownership structures and monetization pressures, posing risks versus its first-mover advantage.

TradingKey - For would-be backers, OpenAI is a bizarre and skewed financial enigma. Indeed, the company was burning cash at record levels even as it generated record levels of revenue. OpenAI's current value will continue to be shaped by this duality and what that might mean for an eventual public offering through an IPO.
Although widespread consumer adoption of ChatGPT is still the biggest driver of OpenAI's rapid growth, the company is crushing early expectations for how well it would perform. OpenAI's growth in annual recurring revenue (ARR) officially smashed through the $20 billion mark in 2025—a mind-blowing year-over-year growth rate of over 233%. This charge was spurred on by a colossal increase in computing power from 0.6 GW to 1.9 GW over the year. OpenAI leadership is on track to further defy belief, with fiscal 2028 revenue guidance of $100 billion and a user base that reached 900 million weekly active users by the close of 2025. This revenue surge notwithstanding, the company still has a hard road to turning a profit, having posted a net loss of $13.5 billion in the first half of 2025 alone as it ramped up infrastructure and R&D spending like never before.
As the fastest-growing consumer app ever, OpenAI’s financials have been a wild ride of enormous revenue and eye-watering losses. The company successfully raised substantial funds in early 2025 to continue its operations, but it still is “burning through cash” faster than ever. OpenAI has projected a net loss of $5 billion in 2024 and is expected to lose as much as $8 billion in 2025.
It is doing so while internal estimates indicate that OpenAI will not be cash flow positive until 2029, and that full profitability or positive free cash flow until 2030 is predicted by, for example, Deutsche Bank. This sustained deficit is being driven by the huge capital requirements for AI research and for AGI-relevant infrastructure. As a result, OpenAI is projecting that its negative free cash flow can reach a staggering size of $143 billion by 2029. To finance this aggressive growth, HSBC analysts think the company could raise, at a minimum, $207 billion in additional capital by 2030.
In contrast, Anthropic, its main rival, is chasing a quicker path to financial sustainability. Anthropic expects to be cash flow positive as early as 2027 and EBITDA positive in 2028. That same year, Anthropic expects to generate annual revenues of $70 billion and a cash flow of $17 billion. While both companies are still in a high-investment, high-loss phase, Anthropic is now poised to become profitable a few years before OpenAI.
The scale of its sustainability is out of proportion. The financial strains built into OpenAI are its massive revenue growth, which doesn’t scale very well. Given this persistent deficit, OpenAI faces some serious strategic challenges as it heads for a potential IPO in 2026. First, the need for constant infusions of cash has pushed the company into convoluted “capped-profit” structures and heavy dependence on partners such as Microsoft and Nvidia, making the ownership picture more complicated for new investors. Second, "cash burn" creates pressure to monetize the platform in ways that may alienate users — like an upcoming plan to test advertisements on ChatGPT. If it can’t make the transition to a higher-margin business model before the market’s taste for AI hype turns, the company could get stuck with a “down round” — or a failed IPO. Investors have to decide if they’re buying a future tech goliath or a company that is forever at risk of eating its own value in chasing AGI.
OpenAI's equity is currently only available to accredited investors and institutional partners, but the public market offers many complicated layers of exposure. Knowing the complexity of the shareholders, synergy partners, and “concept” stocks, you can formulate an “investment thesis” for a basket of stocks that tracks the development of OpenAI until its IPO.
OpenAI's ownership structure has changed significantly as it turned toward a for-profit PBC. Most held (updated): Now selling recommendation. The most held sell-rated stocks among analysts in percentage are:
Microsoft (MSFT) is the largest shareholder and remains the single largest investor, holding 27% in the newly reorganized for-profit entity. Under the profit-sharing arrangement, Microsoft takes 75% of OpenAI’s profits until it recovers its $13 billion investment, then its share of the profits drops to 49%. This means that for all intents and purposes the main public vehicle for OpenAI is Microsoft.
As part of the mission-aligned structure, the nonprofit OpenAI Foundation holds a 26% equity stake (valued at about $130 billion). Importantly, the Foundation retains the right to appoint the board, so development of AGI at the company will continue to be aligned with safety and public benefit.
The rest — 47% — is owned by venture capital and strategic investors, including employee equity and high-profile institutional investors. Thrive Capital — run by Josh Kushner, recently the subject of its own secondary sales — has also played a major role in leading recent secondary sales. Other significant shareholders include SoftBank, which pledged $30 billion in 2025, and Nvidia, which converted its hardware superiority into an equity stake. Notable institutional holders such as T. Rowe Price, Fidelity, and Tiger Global also own substantial slices of the company.
Not many publicly traded companies are very tied to the OpenAI ecosystem — enough that their stock prices move with OpenAI’s technology leaps.
An influential OpenAI partner for OpenAI's custom AI inference chips in late 2025 is Broadcom (AVGO). OpenAI has named Broadcom as its primary engineering partner in this shift, with deliveries expected in the second half of 2026, as the AI giant seeks to diversify away from Nvidia.
Oracle (ORCL) has become a crucial infrastructure provider for OpenAI, signing a multi-year contract to provide the “raw iron” for data centers. Oracle’s cloud infrastructure now fulfills not only Microsoft's but also OpenAI’s colossal training needs, making Oracle a major “infrastructure beneficiary” of OpenAI's rise.
As the primary provider of human conversational data for training of GPT models, Reddit’s (RDDT) data-licensing deal with OpenAI makes it a rare play on the “data supply chain.” Advancements in OpenAI’s models add value to the data Reddit sells.
For retail investors seeking a wide lens on the industry, without having to pick individual stocks, a handful of funds offer access to the private equity world or the wider AI “gold rush.”
If you are hunting for direct private equity access for retail, the Fundrise Innovation Fund enables non-accredited investors to access OpenAI for as little as $10. On the same note, the ARK Venture Fund (administered by Cathie Wood) owns a stake in OpenAI, which lets investors ride the private valuation spikes of the company before it goes public.
In terms of AI ecosystem ETFs, the Alger AI Enablers & Adopters ETF (ALAI) is designed to track the OpenAI ecosystem. The portfolio’s 20 largest holdings list includes Microsoft and Nvidia, but also Broadcom — all three are critical to the OpenAI infrastructure. A few other interesting choices are the Global X Robotics & Artificial Intelligence ETF (BOTZ) and the First Trust Nasdaq Artificial Intelligence ETF (ROBT), which tap the broader trend of productivity gains driven by OpenAI’s technology.
OpenAI’s 2026 IPO horizon: For those with eyes specifically on OpenAI’s 2026 IPO window, Microsoft is once again the best bet, as it holds a direct 27% equity stake and profit share in the entity. But for shareholders fretting about the steep costs of OpenAI’s “cash burn,” Nvidia (NVDA) and Broadcom are a less risky “picks and shovels” trade — they benefit from OpenAI’s astronomical infrastructure spending no matter whether OpenAI itself becomes profitable by 2030. Spreading out some of your capital along these proxies and holding a smaller position in a fund like the Fundrise Innovation Fund will give you the best coverage on OpenAI's pre-IPO path.
In assessing the general-purpose AI landscape, investors have to weigh whether OpenAI’s massive scale and “first-mover” status are dominant enough to dent the smaller, nimbler potential business models of its rivals. While OpenAI is still the highest-valued and reaches the widest audience, the field of competitors has changed dramatically.
The financial statements of OpenAI are both growing rapidly and heavily skewed. By late 2025, the company was said to be in talks for new capital at a potential valuation of as much as $750 billion, though recent secondary market trades in October 2025 pegged the company at around $500 billion. OpenAI’s 2025 ARR is now officially confirmed to be north of $20 billion, putting its price-to-sales (P/S) ratio at around 37.5 based on the target valuation. In contrast, its nearest rival, Anthropic, is seeking funds at a $350 billion valuation as of January 2026. While the overall valuation for Anthropic is lower, its projected revenue of $4.5 billion results in a price-to-sales ratio of around 78, indicating that private investors are willing to pay an even higher premium for Anthropic’s growth relative to its current revenue.
And the time to profits and thus a real business model is a big enough differentiator for investors. OpenAI was “burning cash” at a staggering rate and was not projecting positive free cash flow or profitability until 2030. In contrast, Anthropic, which was founded by former OpenAI co-founders, anticipates becoming free cash flow positive in 2027, and profitable for the first time in 2028. Other players, such as Elon Musk’s xAI or open-source ones like from Meta, have their own trade-offs — xAI is deeply integrated with the X (formerly Twitter) data universe, while Meta’s Llama models are breaking the “closed” capitalist monetization models of both OpenAI and Anthropic by providing outrageously powerful tools for free.
But OpenAI has the virtue of having a broad set of applications and users. ChatGPT remains the top AI consumer app with 900 million weekly users as of December 2025. OpenAI’s offerings are not restricted to text — Sora (video), DALL·E 3 (images), and Whisper (audio) deliver a fuller “multimodal” experience than many competitors. To be clear, one might argue that Anthropic's Claude is technically the best, particularly for code tools like Claude Code, but it doesn't have the cultural impact or potential for mass advertising revenue ($25 billion by 2030) that OpenAI is building up around.
So, is OpenAI the best IPO target?
If OpenAI and Anthropic both were to go public in 2026, OpenAI remains the more attractive “trophy” asset on the strength of its size and the behemoth nature of its growth. Still, for investors concerned about multiples or who want a quicker path to profitability, Anthropic could be considered the more financially conservative play. In the end, OpenAI is the dominant general-purpose AI company today, but the question is whether the investor is more sensitive to OpenAI's market-defining “first-mover” advantage or to the better, safety-centric scaling at Anthropic.