OpenAI has raised its projected spending on cloud and computing infrastructure to $750 billion through 2030, up from a previous estimate of $600 billion. The revised figure, which the company disclosed in recent internal planning documents, signals a major bet on the hardware and data center capacity needed to train and run increasingly powerful AI models.
The $150 billion increase
The new forecast represents a 25% jump over the earlier target. OpenAI now expects to invest $750 billion in cloud services and compute resources over the next six years. That's more than the annual GDP of many countries and underscores the staggering cost of staying competitive in artificial intelligence.
The company did not provide a detailed breakdown of the $150 billion increase. But the spending is tied directly to the massive computing requirements of large language models and the infrastructure to support them.
A $20 billion data center project
Part of that spending includes a $20 billion data center project, one of the largest single infrastructure investments in the tech industry. The project is designed to provide the physical space and power needed for tens of thousands of specialized AI chips.
Data centers of this scale consume enormous amounts of electricity and require advanced cooling systems. OpenAI's project is likely to be located in a region with access to cheap, reliable power, though the company has not publicly confirmed a specific site.
What the spending covers
The $750 billion forecast covers both cloud computing services — rented from providers like Microsoft Azure — and direct capital expenditures on OpenAI's own hardware and facilities. The company has been rapidly expanding its in-house computing capacity to reduce reliance on third-party cloud vendors and to gain more control over its AI training pipeline.
OpenAI's spending plans are among the largest in the tech sector, rivaling the capital budgets of the biggest cloud providers themselves. The company's growth has been fueled by the success of ChatGPT and the broader generative AI boom, which has driven demand for ever-larger models and more training data.
The company has not disclosed a timeline for the $20 billion data center project or how much of the $750 billion will be allocated to it. Those details are expected to emerge as OpenAI finalizes its long-term infrastructure strategy.
