OpenAI Targets $1 Trillion IPO as Annual Revenue Crosses $25 Billion

OpenAI has crossed a revenue milestone that few technology companies reach while still private: $25 billion in annualized recurring revenue, a figure the company confirmed in February 2026 and that has since climbed toward approximately $2 billion per month. The number gives concrete weight to what has been a recurring piece of IPO speculation, and for the first time, the company’s own executives are discussing a public listing timeline with enough specificity to be taken seriously.

The Revenue Story

The path from zero to $25 billion in annual revenue in roughly three years is extraordinary by any measure. OpenAI crossed $1 billion in ARR in late 2023, $5 billion through 2024, and accelerated sharply through 2025 as ChatGPT subscriptions scaled, enterprise contracts multiplied, and API revenue from developers building on the platform became a material revenue stream in its own right.

The March 2026 funding announcement — a $122 billion round that closed on March 31 — disclosed approximately $2 billion in monthly revenue, aligning with the February ARR figure. More than 1,000 customers are now spending over $1 million per year on OpenAI products, a segment that includes major financial institutions, law firms, healthcare systems, and technology companies integrating GPT models into production workflows.

Rival Anthropic, by comparison, is approaching $19 billion in annualized revenue, also growing rapidly after securing a $30 billion raise at a $900 billion-plus valuation in May 2026. The two companies together are defining what a commercially successful frontier AI lab can look like — and establishing financial benchmarks that will set the terms of the IPO conversations that follow.

The IPO Machinery

CFO Sarah Friar has told associates that OpenAI is targeting a regulatory filing in the second half of 2026, with a stock market debut potentially in 2027. The deliberate pace reflects both the complexity of OpenAI’s corporate structure — its conversion from a capped-profit entity to a full for-profit corporation required months of legal and governance restructuring — and a desire to ensure the company’s financials can withstand the scrutiny that public markets apply.

Goldman Sachs, JPMorgan, and Morgan Stanley are in discussions to advise the offering, though no mandates have been formally issued. Bankers involved in those conversations have discussed a potential IPO valuation of up to $1 trillion at listing, though the final number will depend heavily on where interest rates and technology valuations stand when the company actually files.

OpenAI’s current private valuation of $852 billion, established at the March round, already makes it the most valuable startup in history by a significant margin. The gap between its last round valuation and a potential $1 trillion public market target implies relatively modest further appreciation — a $148 billion increase — which bankers close to the deal describe as achievable given the revenue growth trajectory.

The Profitability Problem

The number that complicates the IPO narrative is the burn rate. OpenAI is not profitable and projects annual cash burn reaching $57 billion by 2027. The costs are structural: training frontier AI models at the scale required to maintain competitive advantage consumes enormous compute, the company has made large infrastructure investments in dedicated data center capacity, and a talent base of researchers and engineers competing in one of the tightest labor markets in the technology industry drives personnel costs well above what a traditional software company would spend at comparable revenue.

For context, that level of cash consumption means OpenAI is effectively spending more than $4.5 billion per month more than it earns. At $2 billion in monthly revenue, the implied monthly loss is substantial, and the gap is expected to widen before it narrows as compute and infrastructure costs continue to grow faster than subscription revenue in the near term.

The company’s argument to public market investors will need to address this dynamic head-on. The bull case — the one that underpins the $1 trillion valuation discussion — rests on AI software margins eventually normalizing toward the 60-70% range typical of mature software businesses as training costs decline, inference becomes more efficient, and revenue continues to compound. The bear case is that compute costs remain persistently high, the competitive landscape forces continued heavy investment in research to avoid falling behind, and the public markets prove unwilling to underwrite a company with Amazon-scale capital intensity alongside the growth narrative of a pure software business.

Why This IPO Matters Beyond OpenAI

The OpenAI public listing, when it happens, will be the defining financial event of the current AI wave. It will set a public market reference price for AI capability — establishing whether the premium that private investors have placed on frontier model development is ratified by public investors or discounted.

For the broader startup ecosystem, the outcome has significant downstream effects. Anthropic’s own IPO conversation, which is also active in the background, will be heavily influenced by how OpenAI prices and trades in early public market life. AI infrastructure companies, application-layer startups building on GPT and Claude, and the cloud providers that supply compute to frontier labs will all have their valuations recalibrated once there is a liquid public market reference for the sector.

There is also a governance dimension that public markets will scrutinize in ways private investors have not. OpenAI’s transition from a nonprofit-rooted governance structure to a for-profit corporation has been contentious, involving disputes with former board members, public breakdowns and reconciliations with key figures, and an ongoing negotiation over what residual obligations the company’s original nonprofit mission imposes on its commercial entity. Public company disclosure requirements and shareholder activism will bring a new quality of scrutiny to those questions.

The Competitive Clock

The timing pressure on the IPO is also shaped by competitive dynamics. OpenAI’s revenue lead over Anthropic and Google’s Gemini products is real but not permanent. Google’s I/O 2026 announcements on Tuesday — specifically Gemini 3.5 and the aggressive pricing of Gemini 3.2 Flash at $0.25 per million tokens — signal that Google is willing to deploy its scale advantages to contest the developer market where OpenAI has historically been strongest.

An IPO in 2027 would give OpenAI public capital at a moment when it may need to defend its market position against well-capitalized rivals with lower-cost inference infrastructure and massive distribution advantages. Waiting longer risks either a less favorable valuation environment or a period in which competitive pressure has eroded the revenue growth story that is the central pillar of the listing thesis.

The company that once existed as a research lab dedicated to safely developing artificial intelligence is now a $25 billion revenue business navigating the most consequential IPO decision in the technology industry’s recent history. The numbers are extraordinary. The financial risks are substantial. And the outcome will define what it means to be a public AI company.