The AI giant tripled its top line but burned through $34 billion in a single year, posting a $20.92 billion operating loss
OpenAI pulled in $13.07 billion in revenue during 2025. It also managed to spend $34 billion doing it.
The audited financial figures, which surfaced through tech blogger Ed Zitron and the Financial Times, paint a portrait of a company growing at a blistering pace while hemorrhaging cash at an even faster clip. The operating loss for 2025 came in at $20.92 billion, more than double the $8.78 billion loss recorded in 2024.
Revenue nearly quadrupled year-over-year, jumping from $3.7 billion in 2024 to $13.07 billion. But the cost structure tells a very different story.
The numbers behind the burn #
Total costs ballooned from $12.48 billion in 2024 to $34 billion in 2025. Research and development alone accounted for $19.18 billion of that figure, roughly 56% of all spending.
The company’s expense ratio improved from $2.37 for every dollar of revenue in 2024 to $1.60 per dollar in 2025. Still deeply unprofitable, but the trajectory at least suggests the economics are moving in the right direction.
Microsoft remains deeply embedded in OpenAI’s operations, providing compute infrastructure and participating in revenue-sharing agreements. Without Microsoft’s cloud resources, OpenAI’s cost structure would likely look even more daunting. With them, it means a significant chunk of revenue flows right back to Redmond.
IPO preparations and a massive funding round #
These financials didn’t leak by accident. They’re tied to OpenAI’s preparations for a public offering. The company has filed S-1 paperwork for an IPO expected later in 2026.
Before going public, OpenAI completed a funding round in March 2026 that raised $122 billion, valuing the company at $852 billion post-money.
The revenue jump from $3.7 billion to $13.07 billion represents roughly 253% year-over-year growth.
What this means for investors and the broader AI landscape #
The $19.18 billion R&D line item suggests that staying at the cutting edge of large language models requires spending at a pace that would make pharmaceutical companies blush. Competitors including Google, Anthropic, and Meta are all spending aggressively on their own models.
The Microsoft dependency also deserves scrutiny from prospective public market investors. Revenue-sharing arrangements mean OpenAI’s top line doesn’t fully translate into retained revenue, and reliance on a single infrastructure partner for compute creates concentration risk.
One notable absence from the financial documents: there’s no mention of cryptocurrency assets, tokens, or blockchain-related ventures. Despite the growing intersection between AI and crypto in the broader tech ecosystem, OpenAI appears entirely focused on its core AI development business.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our