OpenAI is unlikely to join SpaceX and Anthropic on the public markets for at least another 12 months, according to PitchBook, which believes the current timeline is too ambitious to complete.
The firm’s senior research analyst Harrison Rolfes warns the company risks being priced against a valuation framework it did not define, on terms set by competitors that scaled with significantly less capital.
Rolfes says the “realistic” IPO window has now shifted from Q4 2026 to mid-to-late 2027.
“OpenAI’s Q4 2026 IPO target was already ambitious given its corporate restructuring. In our view, it is now unrealistic,” he wrote in a report titled OpenAI: The IPO That Cannot Afford to Wait.
He argues that public market investors will require multiple additional quarters of consistent execution before they can “credibly explain” how massive infrastructure commitments translate into sustainable free cash flow, particularly after OpenAI missed several monthly revenue targets earlier this year amid competitive pressure from Anthropic in coding and enterprise markets.
OpenAI currently generates around US$2 billion ($2.8 billion) in monthly revenue, which Rolfes describes as “an extraordinary business,” but highlights growing concern around roughly US$1.15 trillion in infrastructure commitments spanning Oracle, Microsoft, AWS, NVIDIA, AMD, Broadcom and CoreWeave.
“The Oracle contract alone requires US$60 billion a year beginning in 2027, which exceeds OpenAI’s own projected net revenue for that year. These are multiyear commitments that do not flex downward. Revenue does. That asymmetry is what makes the revenue miss dangerous rather than routine.”
He adds that while revenue fluctuations may be temporary, the company’s cost structure is fixed and increasingly central to its IPO trajectory.
“OpenAI’s revenue miss will be forgotten in a quarter. The cost architecture will not. But the less obvious implication is what this does to the sequencing of the entire frontier AI IPO class,” Rolfes said.
The timing, he argues, could reshape the sector’s public market debut entirely. If OpenAI delays further, competitors such as Anthropic and Databricks may define the valuation benchmarks for frontier AI.
“Anthropic and Databricks are both further along in IPO readiness according to our framework. If either company lists first, it will establish the public market comparable for frontier AI on materially cleaner unit economics. OpenAI would then be walking into a valuation framework it did not set, on terms it cannot control. Whichever company prices first defines what “good” looks like in this sector. Being last to market after deploying the most capital is not a positioning any underwriter would choose.”
The comments come shortly after OpenAI CEO Sam Altman and CFO Sarah Friar dismissed reports of internal disagreement over compute spending, calling them “ridiculous.”
“The friction reportedly concerned whether OpenAI could fund its future computing contracts if its growth did not reaccelerate. When the two most senior leaders at a pre-IPO company feel the need to publicly deny a strategic disagreement over capital allocation, the denial itself becomes the data point,” Rolfes argued.
“Friar is asking the right question. The fact that it apparently produced tension tells us how the answer is being received.”
PitchBook also draws a contrast with rivals, noting Anthropic carries significantly lower infrastructure obligations and stronger margins while gaining share in key enterprise segments.
“On an annualised run-rate basis, Anthropic generates approximately US$6 million per employee on 5,000 people. OpenAI generates roughly US$5.6 million per employee on 4,500, with plans to nearly double its head count by year-end. The difference is narrow today, but the trajectory diverges sharply: Anthropic is compounding efficiency as it scales, while OpenAI is adding head count against a cost structure that is already under pressure,” Rolfes said.
In its internal AI business quality framework, PitchBook rates OpenAI’s governance and capital efficiency as key pressure points.
“Governance optionality, already the weakest dimension for OpenAI at 3 out of 10, absorbs the most direct pressure from these recent developments. A public disagreement between the CEO and CFO over spending philosophy during the pre-IPO window is precisely the type of signal this dimension was designed to capture.”
He also warns that revenue quality and capital efficiency metrics are deteriorating as infrastructure commitments expand.
“Holding OpenAI’s 4.2 composite score through the next scoring window looks increasingly unlikely,” he added.
Meanwhile, investor attention is turning to a broader wave of potential US mega-IPOs. According to AUSIEX, companies such as SpaceX, OpenAI and Anthropic are seen as potential blockbuster listings tied to long-term themes including AI, advanced computing infrastructure and space commercialisation.
“For Australian advisers hopeful of access to these larger issuances, any IPO allocations are likely to be closed or tightly constrained, particularly for offshore retail investors. But the potential for a mid-year listing for satellite and space exploration company, SpaceX, raises the prospect that closely watched growth companies could soon become investable through secondary market trading.”
The firm warns that hype around mega-IPOs can distort investor behaviour.
“A disciplined approach to portfolio construction should distinguish between a strong, long-term theme and the short-term emotion that can surround a high-profile listing.
“Access to new opportunities should never just be about the “fear of missing out”, but rather a decision grounded in a clear investment proposition and disciplined portfolio construction.”
It adds that for most offshore investors, participation in IPO allocations is likely to be limited, with secondary market trading remaining the more realistic entry point.
“In most cases, direct participation in an IPO is unlikely to be the main route for offshore retail investors, with allocations typically constrained and access tightly managed. The more realistic pathway for local investors with a case of FOMO is the secondary market after these companies begin trading on a US exchange.”
