OpenAI in 2026: Inside the $852 Billion Company Rewiring the Global Economy
OpenAI in 2026 is no longer just the company behind ChatGPT. With roughly $2 billion in monthly revenue, a $122 billion funding round closed in March 2026 at an $852 billion valuation, and a confidential IPO filing on record with the SEC, it has become the gravitational centre of the largest capital expenditure cycle since the birth of the internet. For investors, the more useful question is no longer “what is OpenAI?” — it is “who earns money because OpenAI exists?”
Key takeaways
- Scale: ChatGPT serves more than 900 million weekly active users; OpenAI reports over 50 million paying consumer subscribers and more than 9 million paying business users.
- Money: revenue reached roughly $13 billion in 2025 and a ~$25 billion annualised run rate by early 2026 — yet projected cash burn of ~$27 billion in 2026 means profitability is not expected before about 2030.
- Structure: in October 2025 OpenAI converted into a public benefit corporation controlled by the nonprofit OpenAI Foundation; Microsoft holds roughly 27% of the company.
- The real trade: OpenAI remains private, but its spending flows directly into listed companies — chipmakers, cloud providers, networking, and energy infrastructure.
From research lab to economic infrastructure
OpenAI was founded in San Francisco in 2015 as a nonprofit research laboratory with a mission to ensure that artificial general intelligence benefits humanity. The economics of frontier AI forced a series of structural changes: a capped-profit subsidiary in 2019, and, in October 2025, a full recapitalisation into OpenAI Group PBC — a public benefit corporation — with the renamed OpenAI Foundation retaining control and roughly a quarter of the equity. That structure is what a future IPO prospectus will be built on.
The product timeline explains the valuation. GPT-3 (2020) proved that scale produces capability. ChatGPT (November 2022) turned that capability into the fastest-adopted consumer product in technology history. The GPT-4 and GPT-5 generations (2023–2026) pushed the models into coding, agents, and enterprise workflows — and turned a research demo into a business generating $2 billion a month.
The Microsoft relationship: from dependency to partnership
Microsoft’s $1 billion investment in 2019 began one of the defining alliances in technology. By 2026 the relationship had been fundamentally rewritten. Under the April 2026 restructuring, Microsoft’s licence to OpenAI’s intellectual property runs to 2032 but is no longer exclusive; OpenAI may now serve customers on any cloud, while committing to purchase around $250 billion of Azure services. OpenAI’s revenue-share payments to Microsoft were capped, cutting tens of billions from the previously projected trajectory.
For investors the key number is ownership: Microsoft holds approximately 27% of OpenAI on a diluted basis, a stake valued around $135 billion — which makes Microsoft stock the closest public-market proxy for OpenAI while the company remains private.
The business model, with real numbers
OpenAI earns revenue from three engines: consumer subscriptions (ChatGPT Plus and Pro), enterprise products (ChatGPT Enterprise, Team, and the Codex coding agent), and usage-based developer APIs, which now process more than 15 billion tokens per minute. The mix is shifting fast: enterprise already contributes more than 40% of revenue and, by the company’s own guidance, is on track to match consumer revenue by the end of 2026.
The uncomfortable half of the income statement matters just as much. Compute costs held gross margins near 33% in 2025, and infrastructure commitments — the $500 billion, 10-gigawatt Stargate programme with SoftBank and Oracle, a ~$300 billion Oracle cloud agreement, a $38 billion multi-year AWS deal — push projected cash burn to roughly $27 billion in 2026 and $63 billion in 2027. OpenAI is simultaneously one of the fastest-growing and one of the most capital-hungry companies ever built.
Competitive landscape: no longer a one-horse race
Consumer mindshare still belongs overwhelmingly to ChatGPT. The enterprise market tells a different story: venture firm Menlo Ventures estimates that Anthropic leads enterprise LLM spending with about 40% share, ahead of OpenAI at 27% and Google at 21%. Meta’s open-weight Llama models and xAI compete for developers, while Chinese open-source models exert continuous pricing pressure on the API business. OpenAI’s moats — brand, distribution, and the largest dedicated compute footprint in private hands — are real, but they are moats that must be refinanced every year.
The investment perspective: follow the capital expenditure
Retail investors cannot buy OpenAI shares yet, and when the IPO arrives — the company filed confidentially with the SEC in mid-2026, reportedly targeting a valuation above $1 trillion, with reports placing the listing window anywhere between late 2026 and 2027 — the entry price will already assume flawless execution. The more durable opportunity has been visible for three years in the supply chain that OpenAI’s spending finances: GPU and accelerator makers (NVIDIA, AMD), custom silicon and networking (Broadcom, Arista Networks), foundries (TSMC), cloud and data-centre landlords (Microsoft, Oracle, Amazon), and the power and cooling infrastructure that 10 gigawatts of AI capacity requires.
A useful discipline: for every headline about a model release, ask which contracts it triggers. A single OpenAI infrastructure announcement in 2025–2026 routinely moved tens of billions of dollars of future revenue onto the books of listed companies.
Orion Capital perspective
Artificial intelligence is becoming a new layer of economic infrastructure, comparable to electricity, the internet, and cloud computing — and infrastructure build-outs reward the suppliers before they reward the builders. At Orion Capital we believe the durable value in this cycle accrues along the full chain: compute, networking, energy, data-centre capacity, and the enterprise software that translates models into productivity. The skill that will matter most this decade is not predicting which model wins, but understanding which cash flows are contractual, which are cyclical, and which are still only promises. OpenAI’s story is extraordinary; the discipline is to invest in the ecosystem’s economics, not its headlines.
Our position, stated so it can be checked later: at the reported IPO ambitions — roughly 35 times forward revenue for a business with ~33% gross margins and no cash flow before 2030 — we would not be buyers at the open. We see better risk-reward in the suppliers holding multi-year contractual revenue from this build-out than in the builder itself. If OpenAI proves us wrong by converting enterprise growth into durable margins, we will say so in an update to this article.
Frequently asked questions
Can I invest in OpenAI directly?
Not yet. OpenAI remains private; a confidential IPO filing was made in mid-2026, but no listing date is confirmed — reports point to a window between late 2026 and 2027. The closest public proxy is Microsoft, which owns roughly 27% of the company.
How much revenue does OpenAI generate?
Roughly $13 billion in 2025, rising to about $2 billion per month by March 2026 — an annualised run rate near $25 billion.
Is OpenAI profitable?
No. Because of compute and infrastructure costs, the company is not projected to become cash-flow positive before approximately 2030.
Who owns OpenAI?
The nonprofit OpenAI Foundation controls the company and holds about 26% of equity; Microsoft holds about 27%; employees and other investors hold the remainder.
Sources and methodology
Figures in this article are drawn from OpenAI’s public disclosures (March 2026 funding announcement), Microsoft SEC filings, and reporting and estimates by Sacra, Bloomberg, CNBC, The Information, and Menlo Ventures, current as of July 2026. Private-company figures are reported or estimated, not audited. This article will be reviewed and updated as new disclosures are published.
Primary sources: OpenAI (openai.com), Microsoft Investor Relations, CNBC, Bloomberg, Sacra and other publicly available reports.
Disclaimer
This report is prepared by Orion Capital for educational and informational purposes only. It does not constitute investment advice, an investment recommendation, or an offer to buy or sell any security. Financial markets involve risk, including the possible loss of capital. Readers should conduct their own research or consult a licensed financial advisor before making investment decisions. The author may hold positions in companies mentioned.
About the Author: Ieva Bružaitė founder of Orion Capital. Read more on the About page.
Related Orion Capital research
- Riviera Capitalism: Why Modern Wealth Looks Different
- Innodata (INOD): The Quiet Architects Behind the AI Revolution
- AMD AI Stock Explosion: The Second Sun of the AI Revolution (2026)