- Google Cloud will provide OpenAI’s infrastructure in a number of countries
- Google also has a cloud deal with OpenAI rival Anthropic
- Then, there are Google’s own in-house AI efforts
OpenAI has now added Google Cloud Platform as a cloud provider for ChatGPT and its APIs in a number of countries, according to CNBC reporting.
This comes as the AI giant diversifies its infrastructure to meet growing demand – the company also famously uses Microsoft, as well as Oracle and CoreWeave.
Once the company’s exclusive cloud provider, Microsoft has lost exclusivity as the relationship between the two companies evolved. Today, Redmond still has first right of refusal, but it’s no longer the sole vendor.
OpenAI adds Google to its list of cloud providers
OpenAI’s use of Google Cloud Platform is a significant win for the company, whose cloud business is much smaller than AWS and Microsoft Azure even though it’s the third-largest cloud provider globally. Its infrastructure will be used in the US, the UK, Japan, the Netherlands, and Norway.
On the other side of the deal, OpenAI isn’t Google’s only AI customer. The tech giant already has a partnership with Anthropic – a rival AI company founded by former OpenAI execs.
OpenAI CEO Sam Altman has previously cited a shortage of GPUs as a motivation to not only increase, but broaden the range of cloud services.
According to Reuters, the deal with Google could have been on the cards for a number of months, but it wasn’t until the exclusive partnerships with Microsoft ended that discussions could progress.
For OpenAI, diversifying suppliers reduces single-vendor lock-in, ultimately giving the company more leverage to negotiate on pricing and to access the latest chips.
However, the deal raises questions about how Google and its rivals will allocate resources, each offering their own alternative to ChatGPT. Looking ahead, OpenAI is on track to finalize accelerator designs soon, which could further reduce its reliance on third parties.
