OpenAI lowers planned compute spending to $600 billion

Six hundred billion dollars—this is the amount that OpenAI now considers sufficient to build the computing infrastructure of its dreams. The paradox is that this represents a reduction in appetite: not long ago, Sam Altman's company's ambitions regarding capital expenditure on compute looked even more grandiose. Yet even the revised figure remains astronomical and exceeds the annual GDP of most countries in the world.

According to data from Chinese publication 36Kr citing sources close to negotiations, OpenAI is updating its long-term capital expenditure plans for investors. The target by 2030 now stands at approximately 600 billion dollars. In parallel, the company is pushing forward a new round of capital raising, the scale of which could exceed 100 billion dollars—which in itself would be the largest private financing deal in the history of the technology industry.

To understand the context of this decision, it is worth recalling how rapidly OpenAI's financial trajectory changed over the past two years. The company, which still generated relatively modest revenue in 2023, by early 2025 reached a level of several billion dollars in annual income. Now OpenAI's forecast suggests that by 2030, cumulative revenue will exceed 280 billion dollars. This means the company expects to enter the top ten largest technology corporations in the world by sales volume—just a few years after its commercial debut.

The structure of expected revenues is particularly noteworthy. According to sources, the consumer and corporate segments of OpenAI's business should contribute approximately equal shares to overall revenue. This is an important signal: the company is no longer banking exclusively on ChatGPT subscriptions for the mass audience. The corporate direction—APIs for developers, custom solutions for business, integrations into corporate systems—is viewed as an equally important growth driver. In fact, OpenAI is building a two-engine model where B2C and B2B segments balance each other, reducing dependence on a single revenue source.

A key change in the updated strategy is linking expenditures to projected revenues. Sources familiar with the course of negotiations emphasize that now investment plans are "more directly linked to expected revenue growth." This might seem like an obvious principle for any company, but for OpenAI it represents a significant shift. Until recently, the logic of Altman and his team was built on the principle of "infrastructure first, monetization later"—large-scale investments in GPU clusters and data centers were viewed as a necessary condition for creating increasingly powerful models, which would then find commercial application. Now investors are being offered more traditional financial logic, where expenditures are justified by concrete revenue projections.

This turn is not coincidental. The venture and strategic capital market, despite sustained enthusiasm regarding AI, is becoming more demanding in justifying investments. When a round of over 100 billion dollars is in question, even the most loyal investors—whether SoftBank, Microsoft, or sovereign funds of the Middle East—want to see a clear link between investments and returns. The reduction of the target spending figure from higher previous benchmarks to 600 billion is essentially a gesture toward financial discipline, a signal that OpenAI is prepared to think not only in terms of technological breakthrough but also in terms of sustainable business.

However, the word "reduction" here is rather conditional. Six hundred billion dollars for computing infrastructure over five years is a sum comparable to the aggregate capital expenditures of all major technology companies in the world over an analogous period. For comparison: Microsoft, OpenAI's largest investor, plans to spend on data centers approximately 80 billion dollars in fiscal year 2025 alone. Google, Amazon, and Meta are also increasing infrastructure investments, but even their combined plans do not reach the scale that OpenAI is indicating.

A natural question arises: where will this money come from? Even if a 100 billion dollar round materializes, this will cover only a small portion of the stated needs. Clearly, OpenAI is counting on a combination of its own growing revenue, partnership investments from Microsoft and other strategic allies, and likely debt financing and infrastructure partnerships with data center operators. It is not excluded that part of these expenditures will be distributed through joint ventures—such as, for example, the Stargate project announced in early 2025.

For the AI industry as a whole, this revision of OpenAI's plans carries a dual signal. On the one hand, the company confirms that the race for computing power continues and investment scales will remain unprecedented. On the other hand, the very fact of downward correction suggests that even the market leader is beginning to align ambitions with reality. The era of unbridled optimism, when any request for capital in the AI sphere was satisfied without additional questions, is gradually giving way to an era where investors want to see a path to profitability. And this, perhaps, is the healthiest signal the market could have received.