PitchBook clients can access this report via the Research Center on the PitchBook Platform.
OpenAI: The scale is real, but the economics are less certain
OpenAI plans to confidentially file its S-1 with the US Securities and Exchange Commission, with Goldman Sachs and Morgan Stanley leading the process toward a Q4 2026 listing near $1 trillion. The reports of OpenAI’s intent to list came just days before Anthropic’s valuation eclipsed OpenAI’s for the first time at $965 billion. Our AI Business Quality framework scores OpenAI at 4.8 out of 10, and a reverse discounted cash flow shows a 3x-7x gap between what OpenAI’s $852 billion valuation implies and what Q1 2026 data supports. We break down the margin paradox, the $38 billion Microsoft cap, the $1.2 trillion compute obligation stack, and the disclosures the S-1 must make before the $1 trillion number can be underwritten.
OpenAI carries the highest price per unit of business quality among the Frontier Five at $177.5 billion per AIBQ point. Its valuation has been decoupled from fundamentals prior to its IPO.
Table of contents
| Key takeaways | 1 |
| Introduction | 2 |
| AI Business Quality (AIBQ) scorecard: OpenAI under the microscope | 2 |
| The revenue engine and the margin paradox | 5 |
| The advertising reality check | 6 |
| The Microsoft reset: Anatomy of a $70 billion to $97 billion swing | 8 |
| Enterprise: The market that OpenAI is losing | 9 |
| The IPO math and the three-way race | 10 |
| What the S-1 must disclose versus what it likely will not | 11 |
| The bottom line and what would change our view | 11 |
| References | 14 |
