According to Techcrunch and blogger Ed Zitron, new internal documents outline how much money is moving between OpenAI and Microsoft. The documents suggest that the cost of simply running OpenAI's models is enormous. Based on these figures, a profitable OpenAI still seems far off.
Citing internal documents, AI critic Ed Zitron says OpenAI paid Microsoft about 493.8 million dollars in 2024 as part of a 20 percent revenue share. Another 865.9 million dollars followed in the first three quarters of 2025. None of these figures have been officially confirmed.
According to Techcrunch, the revenue share works in both directions. Microsoft sends roughly 20 percent of the revenue from Bing and the Azure OpenAI service back to OpenAI. That means the leaked payments represent Microsoft's net share after subtracting the portion it returns to OpenAI.
With a 20 percent cut, the payments imply minimum revenue of about 2.47 billion dollars in 2024 and 4.33 billion dollars in the first three quarters of 2025. Actual revenue is likely higher because Microsoft's payments back to OpenAI are not included. Sam Altman has previously mentioned projecting 20 billion dollars in annualized revenue by the end of 2025.
Zitron also says OpenAI's inference costs—the core compute and operating costs required to run its models—hit 3.77 billion dollars in 2024 and 8.67 billion dollars in the first nine months of 2025. Techcrunch reports that Microsoft credits cover much of the training cost for new models, but inference spending is mostly cash.
| Quarter (calendar year) | Inference cost (US$ bn) | Microsoft revenue share (US$ bn) | Implied OpenAI revenue, minimum* (US$ bn) |
|---|---|---|---|
| Q1 2024 | 0.547 | 0.077 | 0.387 |
| Q2 2024 | 0.748 | 0.110 | 0.548 |
| Q3 2024 | 1.005 | 0.139 | 0.696 |
| Q4 2024 | 1.467 | 0.168 | 0.839 |
| Q1 2025 | 2.075 | 0.206 | 1.032 |
| Q2 2025 | 2.947 | 0.248 | 1.242 |
| Q3 2025 | 3.648 | 0.411 | 2.056 |
Minimum revenue assumes a 20 percent revenue share (implied revenue is about five times Microsoft's portion), via the FT.
Taken together, the data suggests OpenAI is already spending massive sums just to keep its AI models running, far more than many expected.
"Based on these numbers, it appears that OpenAI may be the single-most cash intensive startup of all time, and that the cost of running large language models may not be something that can be supported by revenues. Even if revenues were to match those that had been reported, OpenAI's inference spend on Azure consumes them, and appears to scale linearly above revenue," Zitron writes.
On top of that are training costs, substantial staffing expenses, and other operational spending. Neither OpenAI nor Microsoft has commented on the reports.
