Google Cloud is adding OpenAI models to its list of AI models available to customers, signalling a shift towards cloud providers being the platform on which enterprises manage multiple computing AI models, instead of requiring customers to commit to a ecosystem by a single model provider.
The deal between the two technology bigwigs two days ago means there will be more AI models available to businesses and developers through Google Cloud’s enterprise AI platform.
It will give Google users access to both the cloud giant’s Gemini models, as well as OpenAI models, in addition to the platform’s existing selection of more than 200 first-party, third-party and open models.
According to Google Cloud, customers enjoy more flexibility when developing AI applications and agents. Instead of committing to a single model, customers can select different models according to the cost, capabilities and performance required for individual workloads.
Customers have the option of choosing the right model for any given job, and deliver optimal quality and lower costs, said Thomas Kurian, chief executive officer of Google Cloud.
The collaboration will also make OpenAI’s models and Codex available to developers and businesses already building applications on Google Cloud.
Greg Brockman, president and co-founder of OpenAI, said the partnership would give customers more options for accessing and deploying the company’s AI technology.
“This partnership is an important step in making our most capable models and Codex available where customers are already building, giving them even more choice in how they use frontier AI as we grow the agentic ecosystem together,” he said.
The partnership comes as enterprises increasingly adopt multiple AI models for different use cases rather than relying on a single provider or model.
Research by Enterprise Strategy Group (ESG), in collaboration with IBM, found that 81 per cent of organisations are using three or more generative AI models. This reflects a shift toward fit-for-purpose deployment, where models are selected based on workload needs rather than generalised capability.
