The $300 Billion Gap: OpenAI Takes Investor Calls at $1.2 Trillion While Asking for $1.5 Trillion

Five days after Sam Altman announced that OpenAI would not go public in 2026 — calling this “an ill-advised moment” given the industry’s safety reckoning — the company has quietly opened a different door. According to a Wall Street Journal report published September 15 and confirmed by the Financial Times, investors have approached OpenAI about a new private funding round that would value the company at more than $1.2 trillion, ahead of a listing now expected no earlier than 2027.

Then came the twist. On September 16, the New York Times reported that OpenAI itself believes it is worth considerably more: people familiar with the discussions say the company is targeting a valuation of at least $1.5 trillion. Reuters Breakingviews christened the standoff with a headline that will likely define the deal-talks cycle: “OpenAI plays $1.5 trln chicken with chatbot frenzy.”

The gap between the bid and the ask — roughly $300 billion, larger than the market capitalization of most S&P 500 companies — is now the single most-watched number in private markets.

What we actually know

The reporting converges on a consistent set of facts:

• Investors initiated, not OpenAI. Both the WSJ and FT describe large investors approaching OpenAI with proposals, rather than the company shopping a round. That detail matters: demand-side initiation strengthens OpenAI’s negotiating position and lets the company set the price of admission.
• The anchor is $1.2 trillion-plus. The figure represents a roughly 41% jump from the $852 billion post-money valuation OpenAI set in March 2026, when it closed a round with $122 billion in committed capital — still the largest private raise in corporate history.
• OpenAI wants $1.5 trillion. The NYT reports the company’s internal ask is higher still, grounded in revenue momentum: OpenAI generated more than $40 billion in annualized revenue in August, roughly double its run-rate at the end of 2025. A June S-1 revelation put 2025 revenue above $13 billion against a plan to spend $115 billion over four years.
• Talks are early. No lead investor is named, no round size is fixed, and both papers emphasize that discussions could still dissolve — or, conversely, that the final number could land above $1.2 trillion if competing investors bid each other up.
• The IPO is the backdrop, not the event. Altman confirmed on September 12 that OpenAI will not list in 2026, citing safety timing. Advisers had previously presented the option of waiting until 2027 at a $1 trillion valuation. This round, in effect, lets OpenAI monetize the wait.

Why the step-up is defensible — and contestable

A 41% valuation increase in about six months would be extraordinary for any asset class outside of crypto. But OpenAI’s defenders can point to a business that is compounding faster than almost any in history: revenue is doubling on an annualized basis, ChatGPT’s consumer and enterprise subscriptions continue to scale, and the company has converted its compute commitments into a multi-year pipeline of data-center deals spanning St. David’s Arizona to planned capacity in Ohio.

The skeptics’ case writes itself too. The valuation math — $1.2 trillion against $40 billion of annualized revenue implies a 30x multiple on current sales — prices in years of flawless execution while the company plans to spend more than a hundred billion dollars on infrastructure it must fill. Reuters’ Breakingviews column frames the $1.5 trillion ask as a game of chicken: OpenAI is betting that investors’ fear of missing the defining AI asset of the cycle will force them up to its number, while investors are betting that the IPO delay and the safety turbulence of recent weeks give them leverage to hold at $1.2 trillion.

There is also a competitive screen over the whole negotiation. Anthropic — now valued around $965 billion after its August round, and reportedly preparing its own listing at valuations as high as $2 trillion with Nvidia mulling a $10 billion anchor stake — is circling the same pool of sovereign wealth funds, mega-cap strategics, and private equity dry powder. Every dollar that goes into OpenAI’s round at a higher price is a data point Anthropic’s bankers will use; every dollar that stays on the sidelines is evidence the market has a ceiling.

The private-market flywheel, one more turn

Structurally, this round would extend an unusual equilibrium: the most valuable private company in the world continues raising ever-larger sums in private markets precisely so it does not have to answer to public ones. OpenAI’s dual-entity structure — the nonprofit parent controlling the for-profit arm — was defended by Altman last week as “why” the company has tolerated its complexity: it preserves the option to make decisions, such as slowing a training run or pausing a launch, that are “not obviously in shareholders’ interest.”

For employees and early backers, each private round at a stepped-up valuation provides partial liquidity and a fresh reference price without the disclosure cadence of a public listing. For late-stage investors, it provides a way into an asset that a 2027 IPO will make more expensive — assuming the public market agrees with the private one when the S-1 finally prices. That assumption is doing a lot of work. Private marks have a habit of surviving on scarcity and structure; public markets mark to the quarter.

And the company’s balance sheet gives it the patience to wait for its number. Beyond the $122 billion in committed capital, OpenAI holds a $4.7 billion revolving credit facility, and the March round’s structure — committed capital drawn over time rather than cash upfront — means the war chest accrues on its own schedule. “We don’t feel pressure” was Altman’s phrase about the IPO timeline last week; it applies equally to this round.

What to watch

Three signposts will tell us how the chicken game ends:

1. A lead investor emerging at a number. If a sovereign fund or strategic names itself at $1.2 trillion, OpenAI accepted the market’s bid. If the disclosed price is $1.3 trillion or above, the company won the standoff.
2. Anthropic’s next move. A filed Anthropic S-1 at a $2 trillion target while OpenAI’s round is open would force every allocator in the pool to choose — and would either validate or cap the entire private AI valuation ladder.
3. The revenue print behind the ask. If annualized revenue crosses $45–50 billion by the time terms sheets circulate, the $1.5 trillion ask stops looking like bravado and starts looking like a trailing multiple under 35x — rich, but within the range public investors have paid for scarcer growth stories.

None of this changes the fact that the deal may not happen at all; “early discussions” is doing its customary load-bearing work. But the scenario in which it does close — at any of the numbers now circulating — would make OpenAI the first private company in history to clear a trillion-dollar valuation, roughly two years after it was last valued at $157 billion. Whatever your prior on AI valuations, that trajectory is now the reference case for every venture mark written in the sector.

The gap between $1.2 trillion and $1.5 trillion is not really about $300 billion. It is about who blinks first in setting the price of the defining asset of the cycle — and both sides currently believe the other cannot afford to walk away.