- OpenAI’s ‘Guaranteed Capacity’ allows large companies to unlock three years of guaranteed data processing
- In return, OpenAI gets predictable revenue to fund ongoing data center expansion
- Stargate also introduced new closed-loop water cooling system in a major sustainability leap
ChatGPT maker OpenAI has announced a new ‘Guaranteed Capacity’ plan for companies looking to lock in and reserve long-term access to the company’s computing infrastructure.
This strategic shift comes as AI companies face increased pressure from increasing enterprise AI adoption, which can lead to occasional downtime.
OpenAI recently revealed that around 40% of its $2 billion monthly revenue comes from enterprise customers – a number it expects to continue to grow.
Guaranteed capacity
The company recently revealed that it had 900 million weekly ChatGPT users, but earlier WSJ reporting indicated that OpenAI did not meet internal targets by 2025, including user numbers.
The article also revealed potential tensions between CDO Sarah Friar, who advocates financial discipline, and CEO Sam Altman, who has a ‘buy everything’ approach.
The Guaranteed Capacity program is designed specifically for large-scale applications and agent automation that require continuous and uninterrupted computing, rather than smaller businesses that want to avoid occasional outages.
Eligible businesses can request up to 1B tokens per minute in terms of capacity, and pricing is directly tied to annual usage levels across one-, two-, and three-year plans.
“Customers can draw on this commitment across the portfolio of OpenAI products,” the company said.
Company CEO Sam Altman said in an X post that the update comes in response to customer feedback and will be available “until we sell out of our current allocation to this program.”
We ‘will significantly expand our computing capacity’, says OpenAI
Altman also conceded that predictable revenue from large enterprise contracts and applicants for guaranteed capacity would help the company plan and give it access to more consistent cash to build future data centers.
“Our intent remains to build as many computers as quickly as we can,” Altman added.
OpenAI’s headline-grabber is the Stargate project, which attracted collaboration from Microsoft, Oracle and others. When first announced in January 2025, the company targeted 10 GW of US AI infrastructure by 2029 – a number it had reached just over a year later. In April, the firm noted that it had also recently acquired an additional 3 GW of capacity.
With plans to “significantly expand [its] computing capacity in the coming years,” OpenAI and partners are already exploring future locations across the United States and also note the progress data centers have made in terms of sustainability.
Each building on its flagship Abilene, Texas campus requires the equivalent of two Olympic-sized swimming pools of water, but instead of traditional evaporative cooling towers, it uses a closed-loop system where the water is recirculated through sealed pipes.
“The annual water consumption of the entire cooling system at full deployment is expected to be comparable to a medium-sized office building or about four average households,” the company stated.
By offering companies a way to secure up to three years of dedicated data processing, OpenAI ultimately builds predictability into its revenue to maintain ongoing data center build-out and further improve efficiency.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds.
