AI firm OpenAI fielded some 1.9GW of compute by the end of 2025, the company's chief financial officer, Sarah Friar, has said.

This marked a 9.5× increase from 2023, with an annual increase of 3× – scaling from 200MW in 2023, to 600MW in 2024, and 1.9GW in 2025.

At the same time, OpenAI's revenue increased by 3× every year, and 10× from 2023 – with $2 billion annual recurring revenue in 2023, $6bn in 2024, and more than $20bn in 2025, Friar said in a blog post.

"This is never-before-seen growth at such scale," Friar said. "And we firmly believe that more compute in these periods would have led to faster customer adoption and monetization."

Unmentioned are profits, with the company still running at a heavy loss.

"Compute is the scarcest resource in AI," Friar added. "Three years ago, we relied on a single compute provider. Today, we are working with providers across a diversified ecosystem. That shift gives us resilience and, critically, compute certainty. We can plan, finance, and deploy capacity with confidence in a market where access to compute defines who can scale."

That single provider was Microsoft, an early backer of OpenAI. While OpenAI still plans to spend as much as $250bn on Microsoft Azure services, it has committed to around $1.4 trillion in compute spend across a plethora of cloud services and its own Stargate data center venture.

OpenAI plans to spend $22.4 billion on CoreWeave, has signed a cloud contract with Google that includes the use of its TPUs, and a $38bn multi-year cloud deal with AWS.

This January, it announced a $10bn deal with Cerebras to use its wafer-scale chips.

Last year, OpenAI expanded its contract with Oracle to a record $300 billion over five years – although much of that will be delivered through Stargate.

Alongside an operational data center in Abilene, Texas, which is currently being expanded, Stargates are underway in Michigan, Wisconsin, New Mexico, and Texas.

Outside the US, a project spanning up to 5GW is underway in the United Arab Emirates, a Stargate Norway project with around 100,000 GPUs is being deployed, alongside a smaller 8k GPU Stargate development in the UK.

A 500MW facility has been promised for Argentina, but the project does not appear to have left the letter of intent stage. Similarly, the company has said it is considering developing in Canada.

OpenAI has signed an MoU with SK Telecom to develop an AI data center in the southwest region of South Korea, while OpenAI execs have been spotted touring Asia to scout potential sites in India, Japan, and elsewhere.

Late last year, the company signed an MoU with NextDC for a major Australian data center.

This growing portfolio "turns compute from a fixed constraint into an actively managed portfolio," Friar said.

"We train frontier models on premium hardware when capability matters most. We serve high-volume workloads on lower-cost infrastructure when efficiency matters more than raw scale. Latency drops. Throughput improves. And we can deliver useful intelligence at costs measured in cents per million tokens. That is what makes AI viable for everyday workflows, not just elite use cases."