OpenAI Files for a Trillion-Dollar IPO — The AI Three-Horse Race Is On
Three of the most valuable private companies on the planet are lining up to go public within weeks of each other. SpaceX leads the pack with a roadshow starting this week. Anthropic confidentially filed on June 1. And now OpenAI has joined the queue — targeting a public listing above $1 trillion as early as September 2026.
What we’re looking at isn’t just another batch of IPOs. Combined, SpaceX, Anthropic, and OpenAI could raise more than $200 billion on public markets — more than every US IPO since 2022 put together.
The numbers behind the hype
OpenAI’s confidential S-1 filing (reported by Fortune and the New York Times around May 22) follows a remarkable March 2026 funding round: $122 billion raised at an $852 billion post-money valuation, the largest private fundraise in history. Goldman Sachs and Morgan Stanley are advising on the IPO.
The revenue figures are the most sobering part of any of these three filings. OpenAI generated $13.1 billion in 2025, or roughly $2 billion per month. Enterprise customers now account for more than 40% of that revenue, and OpenAI says enterprise is on track to reach parity with consumer by the end of 2026. That’s a company that started as a research lab in 2015 and is now generating more revenue annually than the GDP of Malta.
Anthropic’s numbers tell a different story. The company confidentially filed its S-1 on June 1 at a $965 billion valuation — higher than OpenAI despite lower revenue. The premium comes from growth velocity: Claude Code, their AI coding assistant, hit $1 billion in annualized revenue by November 2025 (faster than ChatGPT ever did) and reached $2.5 billion by February 2026. Forbes projects Anthropic at around $45 billion in annual revenue by end of 2026.
SpaceX, furthest along in the process, has already made its S-1 public. Ticker: SPCX on Nasdaq. Target valuation: $1.75 to $2 trillion. The company raised $18.7 billion in revenue in 2025. A 21-bank syndicate led by Goldman Sachs is handling underwriting, with a roadshow reportedly beginning June 4 and a target listing date of June 12.
The governance question
What makes OpenAI’s IPO particularly odd is its corporate structure. The company is governed by a 501(c)(3) non-profit holding company with a “capped-profit” subsidiary. The idea — conceived in 2015 — was that AI profits would be capped at roughly 7x the investment, with surplus revenue flowing back into non-profit research. Investors would get their money back plus a reasonable return, and everything beyond that would fund the pursuit of beneficial AGI.
A trillion-dollar IPO complicates that neat narrative considerably. The company has already pivoted to a public benefit corporation structure (after the Sam Altman ouster-and-reinstatement saga of late 2023), but the non-profit board retains control. What “capped profit” means to retail investors buying shares on Nasdaq next year remains one of the vaguest promises in Wall Street history.
Timing and product launches
OpenAI has a habit of timing product releases with capital-raising momentum. GPT-5 launched in September 2025. GPT-5.2-Codex arrived in December. GPT-5.4 came in early 2026. And now GPT-5.6 is expected in June 2026 — a model identifier that briefly surfaced in Codex logs this week, with Polymarket pricing in an 89% chance of a June-30 release.
Whether the model actually ships is another matter. OpenAI has missed or delayed launch dates before, and a model that arrives after the IPO pricing is a model whose capabilities investors have to trust on faith.
The counter-narrative
Not everyone is buying the valuation math. OpenAI’s $852 billion private valuation implies a revenue multiple of roughly 65x annual run-rate. Anthropic’s $965 billion — projected against $45 billion revenue — is closer to 21x. SpaceX at $2 trillion on $18.7 billion revenue is an eye-watering 107x, but investors are pricing in Starship infrastructure, Starlink, and satellite internet, not just launch services.
The bigger concern is cash flow. Sacra reported that revised funding terms force approximately $6 billion in payments this year, pushing OpenAI’s projected burn to $27 billion in 2026 and $63 billion in 2027. The company does not turn cash-flow positive under those projections. Which means that $122 billion in private funding — the largest ever — may not be enough to sustain the burn rate until the IPO actually closes.
Why this matters
I find it interesting that we’ve gone from “AI is a technology experiment” to “three AI-adjacent companies are collectively worth more than the GDP of Japan” in roughly 18 months. The speed of that transition is unprecedented, even for Silicon Valley.
The real question isn’t whether these IPOs will be successful — the demand is undeniable. The question is whether public markets are pricing in reality or narrative. When Anthropic’s valuation exceeds OpenAI’s despite a fifth of the revenue, you’re no longer looking at fundamentals. You’re looking at which growth story the market prefers.
And that’s a game where the house — whichever company closes their IPO first — usually wins.
Sources: Fortune — OpenAI IPO filing analysis, NYT — SpaceX, OpenAI and Anthropic race to go public, CNBC — OpenAI $122B raise at $852B valuation, Reuters — SpaceX accelerates IPO timeline, Forbes — Anthropic valued at $1 trillion, Sacra — OpenAI revenue and burn analysis, SEC.gov — SpaceX S-1 filing
