OpenAI is reportedly setting the stage to utilize Google Cloud’s services to meet its growing computational demands. According to sources familiar with the matter, discussions between the San Francisco-based AI company and the Mountain View tech giant have been ongoing for several months. This collaboration may allow OpenAI to offload some of the server needs for its ChatGPT and Sora platforms to Google. Financial specifics regarding the agreement have not been disclosed, although it is anticipated that Google will provide adequate compensation.
Google Cloud to Reportedly Carry OpenAI’s Compute Burden
A report from Reuters indicates that OpenAI has reached an agreement to leverage Google Cloud’s infrastructure for its chatbot’s computational requirements. The publication, citing three unnamed sources, noted an uptick in OpenAI’s demand for additional cloud servers, prompting the search for this strategic partnership.
Details of the agreement remain sparse, but it is understood that talks between the two companies were held over several months before being finalized in May. The financial aspects of the deal have not been revealed.
This alliance is noteworthy given that both OpenAI and Google are currently vying for dominance in the competitive AI landscape. Both firms have been actively releasing sophisticated large language models, enhancing the functionality of their chatbots, and experimenting with innovative outreach strategies. For example, while Google is integrating its new Gemini model across all platforms, OpenAI has been exploring collaborations with original equipment manufacturers (OEMs).
Interestingly, shortly after Google unveiled its upgraded Gemini 2.5 Pro model, OpenAI launched its own o3-pro model. Following Google’s announcement of reduced application programming interface (API) costs for its latest AI offerings, OpenAI responded by introducing an 80 percent price cut on its o3 model.
Despite the competitive landscape, OpenAI must acknowledge its urgent need to bolster its computing capacity. Following the launch of its image generation feature powered by GPT-4o, company CEO Sam Altman commented that their GPUs were “melting.” Additionally, on Tuesday, ChatGPT experienced several hours of downtime, culminating in a significant outage for users.
