.png){kind=logo}
Record $122 billion financing pushes OpenAI towards a possible public listing, with strategic backers locking in cloud and silicon supply; attention shifts to revenue conversion, margin pressure, and governance disclosure before markets open.
At the opening of the second quarter, OpenAI is finalising a $122 billion financing at a $852 billion post-money valuation, bringing a potential stock market debut into sharper focus and setting a reference point that public investors will inevitably test. Daniel Coventry, Director of Private Equity at Abishai Financial Asia Pte. Ltd., characterises the raise as “a transaction that starts to feel like pre-listing price discovery, because the backers are underwriting access and accountability as much as they are underwriting equity”.
The round draws $116.4 billion of committed capital against a $105 billion target, with headline cheques led by technology groups seeking influence over compute and product roadmaps. Amazon’s $47.7 billion allocation includes a $33.4 billion tranche tied to defined milestones under the financing terms, while Nvidia and SoftBank each commit $28.6 billion as the deal links funding to silicon supply, infrastructure capacity and go-to-market relationships. Coventry’s view is that “capital concentration is not only a financial signal, it is a strategic claim on the next generation of platforms”.
The focus for public market readiness sits on the capacity to turn usage into predictable cash generation. Abishai Financial Asia analysis puts annualised revenue at about $22.1 billion over the latest reporting year, up 233% from about $6.6 billion over the prior year, with a current monthly revenue run-rate of about $2 billion. The platform now supports more than 900 million weekly active users and more than 50 million subscribers, while paying business users currently stand above 9 million. With fewer than 10% of users on paid plans, Coventry argues that “the runway is visible, but investors will demand proof that monetisation expands faster than the cost of serving intelligence at scale”.
Infrastructure defines that cost. OpenAI is projecting compute expenditure of about $664.1 billion through the end of the decade, including planned investment of about $553.4 billion under the Stargate programme to support 10 gigawatts of capacity. Disclosed long-term commitments for data centre capacity include about $276.7 billion with Microsoft and about $42.1 billion with Amazon, putting obligations at more than 14 times the latest annualised revenue run-rate. Coventry frames the scrutiny ahead as “a test of utilisation discipline, because public investors tolerate heavy investment only when the payback is measurable and repeatable”.
Operating metrics already show pressure points. Over the latest fiscal year, inference costs rise about fourfold even as efficiency work targets lower per-token costs, and gross margin tightens from 40% to 33% as operating expenses outpace sales. Over that same reporting year, revenue of about $14.4 billion runs against expenditure of about $8.9 billion, keeping cash flow positive while funding expansion. Coventry describes this as “the moment where the narrative shifts from growth alone to the economics of growth, because the market will ask what happens when the easy capacity gains are gone”.
Governance is the second test. Bank engagement begins earlier in the year, and the company is weighing a listing window running from the final quarter of the year into the opening stretch of the next, depending on market conditions and internal controls. The public benefit corporation structure remains central, with a nonprofit parent holding 26% of equity valued at about $143.9 billion in the current framework, alongside Microsoft at 27%, employee holdings at 26% and other investors at 21%. Secondary market pricing points to valuations approaching $1.1 trillion, implying about 65 times the latest annualised revenue, while projections indicate losses of about $15.5 billion over the coming reporting year and additional funding needs of about $229.1 billion through the end of the decade under one banking model. Coventry’s assessment is that “valuation is not the hard part, governance is, because disclosure and control structures are what determine whether a premium survives in public markets”.
Regulatory and legal exposures also matter. Compliance under the EU AI Act adds operational rigour to model development and deployment, while litigation from co-founder Elon Musk, seeking about $148.3 billion in damages in the current claim, reinforces the need for clarity on accountability and oversight. Abishai Financial Asia continues to track the balance between revenue momentum, infrastructure commitments and governance readiness as OpenAI moves closer to a public market test.
Abishai Financial Asia Pte. Ltd. (UEN: 201016239E) is a Singapore asset manager established in 2010, positioning itself as a research-led partner in capital allocation.
Approach: Risk-aware capital compounding in public markets via active equity selection, bottom-up research, disciplined rebalancing and overlay tools that support resilience and capital efficiency, including systematic tilts, opportunistic hedging and drawdown-aware risk controls.
Governance: Macro-aware risk budgeting with explicit limits, exposure and concentration guardrails, liquidity filters, stress testing, transparent attribution and ongoing monitoring with clear commentary.
Sustainability: ESG integration through sector and issuer assessment, engagement expectations and governance screens, embedded where financially material across the investment lifecycle.
Access: The firm is exploring compliant product wrappers and distribution pathways that could, subject to suitability criteria, broaden selected solutions to retail-qualified investors over time.
Further information: https://abishai.com
Media contact: Peng Joon, p.joon@abishai.com
