OpenAI announced that it will use Google's cloud infrastructure for ChatGPT and its API. This decision was made due to a lack of capacity and a growing number of users. Previously, OpenAI relied exclusively on Microsoft, but later the companies began to position each other as competitors, writes CNBC.
OpenAI now has several cloud providers: Google, Microsoft, CoreWeave, and Oracle. Google’s services will operate in the United States, Japan, the Netherlands, Norway, and the United Kingdom. The move is significant for Google, as its cloud division is smaller than similar services from Amazon and Microsoft. Google is also working with Anthropic, a company founded by former OpenAI employees.
Back in 2023, Oracle announced a partnership with Microsoft and OpenAI to extend the Azure platform to Oracle Cloud Infrastructure. In March 2024, OpenAI signed a five-year deal with CoreWeave worth nearly $12 billion.
In January, Microsoft agreed to a new model of cooperation with OpenAI: it now has the right of first refusal if OpenAI needs new resources, but is not the exclusive supplier. At the same time, Microsoft retains exclusive rights to OpenAI's APIs.
In April, OpenAI CEO Sam Altman confirmed that the company was facing a capacity shortage and called for GPU sharing.