OpenAI has confirmed it is the company behind Oracle’s massive $30 billion per year cloud services deal disclosed last month. The Wall Street Journal first reported the figure, and OpenAI CEO Sam Altman later acknowledged the partnership, though without confirming the exact dollar amount, in an X post and a company blog update.
The deal provides OpenAI with 4.5 gigawatts of capacity from Oracle as part of Stargate, the $500 billion data center project announced in January by OpenAI, Oracle, and SoftBank. The capacity involved is enormous, equivalent to the power output of two Hoover Dams and enough to supply around four million homes.
The new infrastructure will be built at the Stargate I site in Abilene, Texas, significantly expanding OpenAI’s computing capabilities for training and deploying its advanced AI models. However, this undertaking comes with huge financial and energy costs for both companies. Oracle invested $21.2 billion in capital expenditures last fiscal year and plans to spend another $25 billion this year, primarily on data centers, to meet rising demand from OpenAI and its existing customers.
The scale of the OpenAI deal is striking, as Oracle’s entire cloud business generated $24.5 billion in revenue during its last fiscal year across all customers. The agreement also comes as OpenAI’s annual recurring revenue recently reached $10 billion, nearly double last year’s figure, but still far below the cost of this single commitment.
This partnership highlights the enormous infrastructure demands of next-generation AI development and positions Oracle as a major player in the race to power artificial intelligence at scale.
