TechnologyNewsDeveloping story

OpenAI reportedly plans to buy Google AI chips as it expands computing capacity

Bloomberg and Reuters report a planned chip purchase, but final terms are unknown. The proposal would add to OpenAI’s existing use of Google chips and Nvidia hardware.

OpenAI logo seen through a handheld magnifying glass
A magnifying glass is held over an OpenAI logo on a screen in 2023. File photograph. OpenAI logo with magnifying glass (52916339167), by Jernej Furman from Slovenia (resized and converted to WebP). CC BY 2.0.Image sourceImage licence: CC BY 2.0
WhatsAppXLinkedInEmail
Save for later

OpenAI intends to buy AI chips from Google, Bloomberg reported on October 28, citing a person familiar with private discussions. The proposed purchase could give the ChatGPT maker another source of computing power for its AI build-out, but final terms are unknown. Reuters separately reported the plan, citing a report by The Information. OpenAI and Google declined to comment to Bloomberg.

How Google chips fit OpenAI’s computing plans

A purchase would extend an existing relationship. Reuters reported in June 2025 that OpenAI had begun renting Google’s AI chips. OpenAI said the previous week that it would make its frontier AI models available through Google Cloud, Reuters reported. That arrangement to offer models through Google’s cloud is separate from the reported chip purchase.

OpenAI is already among the largest buyers of Nvidia graphics processing units, which it uses to train and deploy AI models, according to Reuters. Bloomberg reported that OpenAI also plans to start using its own chips later in 2026, including Jalapeño chips developed with Broadcom. The reports did not establish how many Google chips OpenAI might buy or when a purchase might take effect.

Google’s tensor processing units, or TPUs, were first developed for search and have since become an alternative to Nvidia GPUs for AI work, Bloomberg reported. Reuters said Google began selling TPUs directly to some customers in 2026. Buying chips would give OpenAI a different route to the hardware it has previously rented, alongside its Nvidia purchases and planned in-house chips.

Oracle adds Google TPUs to its cloud lineup

Oracle announced on October 27 that it would make Google TPUs, OpenAI’s Jalapeño chip and Cerebras hardware accessible through Oracle Cloud Infrastructure, CNBC reported. The announcement concerns hardware offered to Oracle cloud customers, a separate development from the purchase under discussion between OpenAI and Google.

CNBC reported that Oracle’s offering would be the first time Google’s custom chips ran on another major cloud provider’s infrastructure. Oracle executive Mahesh Thiagarajan said chip scarcity remained a problem and that customers such as OpenAI could benefit from a choice of accelerators. The cloud expansion illustrates another way Google’s hardware could reach AI developers as OpenAI seeks more computing capacity.

Read in depth: Technology

Sources and context

AI-assisted article checked against the listed sources. NewsJaws did not conduct interviews or attend the reported events.

Topics in this story

About NewsJaws Desk

AI-assisted reporting and explainers reviewed against the linked source documents. No claim of on-scene reporting or original interviews.

More from NewsJaws Desk

More from Technology

See all