Model

OpenAI seeks $1.4 trillion the same day it cut the price of intelligence fivefold

A reported $30 billion pre-IPO round would mark OpenAI about 64% above its March valuation with no listing date attached, on the same day it shipped near-frontier intelligence at a fifth of Astra's price. The CEO deferring the IPO over a "10% chance of killing everybody" is asking for more time and more money.

Vincent JiangVincent Jiangthree-minute read
Share
Sam Altman seated on stage at a TED conference, looking to the side, wearing a grey sweater.
1 / 7Slide 1 of 7
OpenAI CEO Sam Altman on stage at a TED conference. He has ruled out a 2026 IPO even as OpenAI seeks a reported $30 billion at a $1.4 trillion valuation.

One stage, two price moves

At DevDay in San Francisco on Tuesday, Sam Altman sold developers GPT-6.1 Sol: near-Astra intelligence at one-fifth of Astra's token prices, $2 per million input tokens and $0.10 cached 1. A different price landed the same morning. OpenAI$1.18T — OpenAI, private, latest valuation $1.18T is in talks to raise at least $30 billion at a valuation near $1.4 trillion, excluding the new money, after pushing its listing to next year 234.

March's paper is still waiting

The last private raise, $122 billion in March, closed at an $852 billion valuation with equity led by Nvidia−2.82% — Nvidia, down 2.82 percent today, Amazon+3.66% — Amazon, up 3.66 percent today and SoftBank, and was framed as the money that would carry OpenAI to a listing 56. Altman has since called 2026 "an ill-advised moment" for an IPO: "I think it is unacceptable to be taking like a 10% chance of killing everybody by the end of the decade" 4. The new round is a bridge in place of the listing 23.

A 64% markup with no exit date

At $1.4 trillion, the ask sits roughly 64% above March's mark, on a line that has already run from $157 billion in October 2024 to $500 billion a year later 67. PitchBook analyst Harrison Rolfes calls it "a capital raise, not price discovery": OpenAI spent about $34 billion last year, and nothing is priced yet 6. Secondary trades in September implied closer to $475 billion, a 44% discount to the primary mark 7.

From $157 billion to a possible $1.4 trillion in 24 months

  • Primary valuation
  • Estimate
$0B$500B$1,000B$1,500BOct 2024Oct 2025Mar 2026Sep 2026 (talks)$1,400Bsold as the last private raise
Data
Primary valuation
Oct 2024$157B
Oct 2025$500B
Mar 2026$852B
Sep 2026 (talks) (estimate)$1,400B
Primary valuation marks in US$ billions. The September 2026 point is a reported target in early talks, not a closed round.2,5,6,7

The unit price of intelligence is falling

Sol is the only new model OpenAI was willing to ship this week: GPT-6.1 Astra stays locked, pulled Monday for missing the company's safety bar 89. Sol finishes a scientific task for an average $5.47, against $23.80 for Astra and $23.21 for Anthropic's$350B — Anthropic, private, latest valuation $350B Opus 5.5 1. The cuts are competitive, not charity: OpenAI has been discounting to "compete for cost-conscious customers against Anthropic and a growing list of Chinese rivals" 10.

Near-frontier science for under a quarter of Astra's cost per task

$0$5$10$15$20$25GPT-6 Astra$23.8Claude Opus 5.5$23.21GPT-6.1 Sol$5.47shipped the day the $1.4T ask landed
Data
Value
GPT-6 Astra$23.8
Claude Opus 5.5$23.21
GPT-6.1 Sol$5.47
Average cost per task on Terminal-Bench Science 0.1 at maximum reasoning effort, US dollars, per OpenAI's published benchmark results.1

Revenue is outrunning the cuts

Deflation has not reached the top line yet: annualized revenue ran near $40 billion in August, roughly double the end of 2025 10, and about $68 billion by Tuesday, up more than 70% since the quarter began, with enterprise sales more than doubling since July 811. The audited 2025 ledger still shows a $38.5 billion net loss on $13.07 billion of revenue 11. Oracle−1.00% — Oracle, down 1.00 percent today, the compute partner, rose 5.3% on the readout 11.

Watch the term sheet, then Anthropic

CFO Sarah Friar declined to discuss the talks and called OpenAI "very well capitalized"; Altman says he has no timeline for a listing 8. Anthropic has filed confidentially and could list this fall 3. When the first lab prices publicly, every private mark inherits a judge.

Get Model by email, free

The AI labs, their models, and the business behind them. In our daily email, every weekday.

Loading NVDA

Reader comments

0 comments

    Model

    HeyGen's Avatar 6 scored 84% on realism, but its 42.2% pick rate was below chance

    Avatar 6 launched November 17, 2026, with an 84% realism figure and first place in a blind arena; the score doubles a below-chance pick rate, and the arena capped 97% of clips at 35 seconds. That is the evidence base under a company claiming $200 million of annual revenue.

    Model

    MiniMax hit a record low, about 90% below its January peak, as a US rule gates its rented Nvidia compute

    Washington's new remote-access rule puts a US license between China's listed model makers and the offshore Nvidia capacity they rent. MiniMax, the purest listed bet on Chinese foundation models, holds the price tag.

    Model

    OpenAI cuts flagship pricing 20% to match Anthropic as heavy users call GPT-6.2 Astra unusable

    OpenAI cut its flagship's price to match Claude Fable 5.2's rate, and the users who run models hardest rejected it, 24 hours before Anthropic priced a $1.4 trillion IPO on the assumption the quality gap holds. The independent labs say the gap is real, for now.

    Model

    Microsoft books Azure revenue on Chinese open models running on Baseten hardware it does not own

    Microsoft now bills customers' committed Azure dollars for open models from Chinese labs running on Baseten's hardware, a toll booth earning rent on compute it does not own.

    Model

    Alibaba stops giving away its best AI model, three days before earnings

    Alibaba launched Qwen4-Max on November 16 and kept the 2.4-trillion-parameter flagship behind its own paid API, promising open weights only for smaller models. Thursday's report is the first test of whether Qwen's three billion downloads become cloud revenue, and it lands two days after a US rule put the chips behind the model under license.

    Model

    Perplexity claims $1 billion of revenue on five times less spend than Anthropic's $1.4 trillion IPO

    Perplexity says it hit $1 billion in recurring revenue on a fraction of Anthropic's spend, three days before Anthropic prices a $1.4 trillion IPO on the opposite assumption. Whose economics hold will only be settled by a filing Srinivas now says could come in early 2027.

    Subscribe to TheInference for free