In the dynamic and often opaque world of Artificial Intelligence, OpenAI has emerged as a colossus, capturing headlines and imaginations alike. While the promise of groundbreaking AI permeates every discussion, a closer look at the company’s financial underpinnings reveals a more intricate, and perhaps precarious, reality. Recent leaks, meticulously analyzed by tech blogger Ed Zitron, offer an unprecedented peek behind the curtain of OpenAI’s financials, shedding light on its revenue streams and, crucially, its colossal compute costs over the past couple of years.
The Microsoft Connection: A Mutually Beneficial, Yet Complex, Alliance
At the heart of OpenAI’s financial narrative lies its deep-seated partnership with tech giant Microsoft. This alliance, bolstered by a staggering $13 billion investment from Microsoft, has been instrumental in OpenAI’s rapid ascent. As part of this arrangement, OpenAI shares a portion of its revenue with Microsoft. While neither company has publicly confirmed the exact percentage, it’s widely reported to be around 20%.
The leaked documents provide concrete figures: in 2024, Microsoft reportedly received $493.8 million in revenue share payments from OpenAI. This figure saw a substantial jump in the first three quarters of 2025, reaching an impressive $865.8 million. Based on the rumored 20% revenue share, these numbers suggest OpenAI’s revenue was at least $2.5 billion in 2024 and approximately $4.33 billion for the first nine months of 2025. These figures align with, and in some cases, even exceed previous estimates from publications like The Information, which placed OpenAI’s 2024 revenue around $4 billion and its first-half 2025 revenue at $4.3 billion.
However, the relationship is a two-way street. Microsoft doesn’t just receive; it also gives back. A source familiar with the matter revealed to TechCrunch that Microsoft shares a portion of its revenue generated from Bing and its Azure OpenAI Service with OpenAI. This reciprocation is estimated to be around 20% of the revenues from these specific offerings. The Azure OpenAI Service, in particular, is a crucial revenue driver, providing businesses and developers with cloud access to OpenAI’s powerful AI models.
It’s important to note that the leaked payments from OpenAI to Microsoft are understood to be based on net revenue share, meaning they exclude any payments Microsoft makes back to OpenAI for Bing and Azure royalties. This nuance adds another layer of complexity to understanding the true financial flow between the two entities.
Beyond Revenue: The Mammoth Cost of Running AI
While revenue figures are crucial, they represent only one side of the financial equation. The other, and perhaps more daunting, side is the sheer cost of operating and scaling advanced AI models. This is where OpenAI’s expenditure on compute power becomes a central focus.
Inference, the process of using a trained AI model to generate responses, is computationally intensive and therefore expensive. According to Ed Zitron’s analysis of the leaked documents, OpenAI may have incurred approximately $3.8 billion in inference costs in 2024. This figure surged dramatically to an estimated $8.65 billion in just the first nine months of 2025.
Historically, OpenAI has leaned heavily on Microsoft Azure for its compute needs. However, recognizing the need for diversification and potentially better terms, the company has also forged partnerships with other cloud providers like CoreWeave and Oracle, and more recently, with AWS and Google Cloud. Despite these efforts, the bulk of its inference compute likely still originates from its primary partner, Microsoft.
Previous reports have estimated OpenAI’s total compute spend for 2024 at around $5.6 billion, with a ‘cost of revenue’ of $2.5 billion for the first half of 2025. While training costs are often offset by non-cash credits from Microsoft’s investment, inference costs represent a significant cash outlay for OpenAI.
A Troubling Implication: Spending More Than Earning?
The stark contrast between OpenAI’s burgeoning revenue and its escalating compute expenditure raises a critical question: could the company be spending more on running its AI models than it is generating in revenue? The figures, as they stand, strongly suggest this may be the case.
This potential financial strain has significant implications for the broader AI industry. OpenAI, as a leader in the private AI market, is often seen as a bellwether for the sector’s health. If even this titan is grappling with profitability, it could cast a shadow over the astronomical valuations and investments being poured into other AI startups.
The narrative of a potentially unprofitable AI giant fuels the ongoing "AI bubble" chatter, a concern that has permeated discussions from Silicon Valley boardrooms to Main Street. The fundamental question arises: if the cutting edge of AI development is financially unsustainable in its current form, what does this portend for the future of AI investment and innovation?
Looking Ahead: The Continued Quest for Profitability
OpenAI’s CEO, Sam Altman, has painted an ambitious picture of the company’s future, projecting annualized revenue run rates well above current reports, potentially reaching $20 billion by the end of the year and even $100 billion by 2027. These are projections, of course, and not concrete guarantees of actual revenue. However, they signal a strong belief in continued exponential growth and the increasing demand for advanced AI capabilities.
The company’s ability to translate its groundbreaking research and powerful models into sustained profitability will be a defining factor in its long-term success and its impact on the AI landscape. The current financial disclosures, while incomplete, offer a valuable, albeit sobering, perspective on the immense capital required to fuel the AI revolution. As the industry matures, the focus will inevitably shift from pure innovation to sustainable business models, and OpenAI’s financial journey will undoubtedly be a closely watched case study.
OpenAI declined to comment on these financial revelations, and Microsoft did not respond to requests for comment. The exact details of their financial dance remain, for now, a closely guarded secret, but the leaked figures provide a compelling glimpse into the high-stakes game of building the future of intelligence.
Join the Conversation: What Does This Mean for AI’s Future?
These insights into OpenAI’s finances spark vital conversations about the economic realities of advanced AI. Are we witnessing the dawn of a new era, or are we caught in an unsustainable financial frenzy? Your thoughts are welcome in the comments below. Let’s discuss the implications for developers, businesses, and the future of technology.