Financial scrutiny of OpenAI is intensifying as leaked documents reveal details about the company’s revenue and substantial compute costs. The information, obtained by tech blogger Ed Zitron, provides a rare glimpse into the financials of the artificial intelligence leader amid ongoing discussions about a potential initial public offering (IPO). These disclosures are fueling debate about the economic viability of large language models and the broader AI landscape.
The leaked data indicates Microsoft has made significant revenue share payments to OpenAI. In 2024, Microsoft paid $493.8 million, a figure that rose to $865.8 million in the first three quarters of 2025. This revenue sharing is tied to Microsoft’s over $13 billion investment in OpenAI.
Understanding OpenAI’s Revenue Streams and Costs
The financial relationship between OpenAI and Microsoft is complex. While Microsoft shares 20% of OpenAI’s revenue, it also receives revenue from services powered by OpenAI’s technology, such as Bing and Azure OpenAI Service. A source familiar with the matter confirmed to TechCrunch that the reported payments to OpenAI represent Microsoft’s *net* revenue share, meaning royalties from Bing and Azure are deducted before calculating the amount paid.
Estimating Microsoft’s total contribution is difficult, as the tech giant does not publicly disclose revenue figures for Bing or Azure OpenAI. However, based on the 20% revenue share, OpenAI’s revenue is estimated to have been at least $2.5 billion in 2024 and $4.33 billion in the first three quarters of 2025. These figures align with previous reports from The Information, which estimated OpenAI’s 2024 revenue at around $4 billion and its revenue for the first half of 2025 at $4.3 billion.
OpenAI CEO Sam Altman has stated the company’s revenue is “well more” than $13 billion annually, projecting an annualized revenue run rate exceeding $20 billion this year and potentially reaching $100 billion by 2027. These projections, however, are not formal guidance.
The Rising Cost of AI Inference
Alongside revenue, the leaked documents highlight OpenAI’s escalating compute costs. The company reportedly spent approximately $3.8 billion on inference in 2024, increasing to $8.65 billion in the first nine months of 2025. Inference refers to the computational power required to run trained AI models and generate responses.
OpenAI has historically relied heavily on Microsoft Azure for compute access, but has recently diversified its partnerships to include CoreWeave, Oracle, AWS, and Google Cloud. While training AI models involves significant upfront costs, a source indicated that OpenAI’s inference costs are primarily cash expenses, unlike the credit-based training costs provided by Microsoft.
These numbers suggest OpenAI may be spending more on inference than it is currently earning in revenue, raising concerns about the company’s profitability. This situation is prompting wider discussion about the financial sustainability of AI development.
Implications for the AI Industry
The potential for OpenAI to be operating at a loss despite substantial revenue raises questions about the broader artificial intelligence market. If a leading company like OpenAI struggles with profitability, it could signal challenges for other AI startups and the massive investments flowing into the sector. The situation fuels ongoing debate about the “AI bubble” and the realistic timelines for achieving returns on investment in this rapidly evolving field.
The high cost of machine learning infrastructure is a key factor. Maintaining and scaling the computational resources needed to power large language models is expensive, and companies must find ways to optimize costs or generate sufficient revenue to offset them. This is driving innovation in areas like model compression and efficient hardware.
Additionally, the complex revenue-sharing agreements between OpenAI and Microsoft demonstrate the intricate financial arrangements that characterize the AI ecosystem. Understanding these relationships is crucial for assessing the true economic value of AI technologies.
OpenAI and Microsoft both declined to comment on the leaked documents. The information underscores the need for greater transparency in the financial performance of AI companies.
Looking ahead, the next few months will be critical as OpenAI continues to refine its business model and explore new revenue streams. Investors and industry observers will be closely watching for further financial disclosures and any indications of a path to sustained profitability. The timing and details of a potential IPO will also be a key focus, as will the company’s ability to manage its escalating compute costs and maintain its competitive edge in the rapidly evolving AI landscape.
