OpenAI will make its frontier artificial intelligence models and Codex coding agent available through Google Cloud in the coming weeks, a move that deepens ties between two fierce rivals and highlights how major cloud providers have become the primary distribution channels for top AI models. The partnership brings OpenAI’s latest models, including GPT-6 Astra, to Google Cloud’s Model Garden, allowing customers to purchase them through Google Cloud Marketplace and apply existing cloud spending commitments toward the purchases.
For enterprise buyers, the arrangement removes a layer of friction. Companies already invested in Google Cloud can now access OpenAI’s models without renegotiating contracts or shifting budgets, which matters for large organizations managing sprawling procurement processes. Customers in regulated industries gain access through Google Cloud’s access controls, data governance tools, and infrastructure, enabling them to build agentic AI workflows within a framework that satisfies compliance requirements.
Google has become the latest hyperscaler to offer OpenAI models after OpenAI’s exclusive distribution arrangement with Microsoft expired in April. Since then, OpenAI has expanded distribution to Amazon Web Services and now Google Cloud, each a major player vying for AI market share. The deal adds another layer to an increasingly complicated relationship between OpenAI and Google, which compete directly in core AI products and search even as the economics of building and deploying large models push them toward cooperation.
Google has positioned its cloud unit as a neutral platform offering customers broader choice, a strategy that took shape after it struck a cloud deal with OpenAI to provide computing capacity last year. Thomas Kurian, chief executive of Google Cloud, has previously described the platform as open and welcoming to any model popular with developers. Adding OpenAI’s models strengthens that positioning rather than undermining it, because customers increasingly want options rather than lock-in.
The arrangement reflects a broader industry shift in which hyperscalers are becoming the main commercial channels for frontier AI labs. Cloud providers typically take a cut from sales through their platforms, creating a revenue stream that rewards them for hosting competitors’ models. Anthropic, an OpenAI rival backed by Amazon and Google, has seen 47 percent of its 2025 revenue come from cloud providers, according to its IPO filing. That figure illustrates how dependent AI labs have become on cloud infrastructure partners, not just for computing power but for reaching customers at scale.
What makes this partnership notable is the tension at its core. OpenAI and Google are competitors in the most direct sense, battling for dominance in AI models and search. Yet the cost and complexity of building and deploying large models have made cooperation unavoidable. Google gets a cut of OpenAI sales and a stronger claim to being the neutral platform of choice. OpenAI gets access to Google’s enterprise customer base and the credibility that comes with operating inside a major cloud provider’s governance framework.
Whether this balancing act holds is an open question. Cloud providers are simultaneously partners and competitors to the AI labs they host, and the lines will only blur further as each side expands its ambitions. For enterprise customers, the immediate effect is more choice and simpler purchasing. For the industry, it is another sign that the AI race is no longer just about who builds the best model, but about who controls the channels through which those models reach the world.
