OpenAI announced on October 21, 2026 that its most advanced models and Codex are coming to Google Cloud's Model Garden catalog, with general availability planned in the coming weeks. It mainly matters to companies already on Google Cloud: they'll be able to pick OpenAI models without leaving that platform.
What we know
- The announcement: OpenAI says it will bring its "frontier models" and Codex agentic workflows to Google Cloud's Model Garden, the catalog where the platform gathers its own and third-party models.
- How to buy: per OpenAI, through Google Cloud Marketplace. Enterprises with Google Cloud spend commitments will be able to draw down those commitments.
- When: OpenAI says general availability is planned "in the coming weeks." In the meantime, there's a form on its site to request access.
- More Google products: according to the announcement, OpenAI models will "soon" be available in Google Antigravity and the Gemini Agent, with no date.
- Data and control: OpenAI says its models will be paired with Google Cloud's built-in access controls and centralized data governance, and frames this as suited to regulated industries too.
- What they said: Greg Brockman of OpenAI talks about making its models and Codex available "where customers are already building." Thomas Kurian of Google Cloud talks about offering a broad selection of models.
- Microsoft: the announcement notes that Microsoft remains OpenAI's primary cloud partner.
What changes and what doesn't
What changes is where OpenAI's models can be bought: from Google Cloud as well as from OpenAI's own platform. Nothing changes today, since general availability hasn't arrived. The announcement doesn't name which models will be offered, doesn't publish pricing or regions, and says nothing about ChatGPT or the OpenAI API, which stay as they were.
How to tell if it affects you
- If your business doesn't use Google Cloud, this changes nothing for you. The same goes if you use ChatGPT or the OpenAI API directly.
- If you already use Google Cloud for AI, for example a website chatbot or an automation built on Gemini models, you'll be able to try OpenAI models on the same platform and the same bill once general availability arrives.
- If you have a Google Cloud spend commitment, ask whoever manages your account whether it makes sense to charge these purchases against it: OpenAI says you'll be able to.
- Before switching models, test with your own real questions and compare answers, latency and cost side by side. With no published pricing, the math isn't possible yet.
- Check what data you send. If the chatbot or automation handles customer data, confirm which access and data controls apply to the chosen model inside your Google Cloud project.
Related: Gemini Agent: Google Gives AI Its Own Workspace Account
Sources
- OpenAI, "OpenAI models are coming to Google Cloud", October 21, 2026
- OpenAI Developer Community, announcement, October 21, 2026
Updates: once OpenAI or Google Cloud confirm models, pricing and general availability, it will be noted here with a link.