OpenAI is putting its AI models on Google Cloud. The company announced Wednesday that its frontier models — GPT-Astra, GPT-6.1 Sol, and GPT-6 Luna — plus its Codex coding agent will land in Google Cloud’s Model Garden and Marketplace. General availability is expected in the coming weeks, though select enterprise customers are getting private-preview access starting today.

What’s actually changing

Google Cloud customers will be able to buy OpenAI models directly through the Marketplace and count the spend against their existing cloud commitments. No separate procurement dance, no second billing relationship. The models deploy from Model Garden, Google’s catalog for third-party and first-party models, and all usage lands on a single Google Cloud invoice that also draws down Flexible Savings Plans.

That’s a meaningful detail. It means a team running on Google infrastructure can build agents that mix OpenAI’s models with Google’s own Gemini 4 Argon, which launched in September, without leaving the Google stack.

The lineup is tiered by workload. GPT-Astra is the flagship, aimed at demanding work across coding, computer use, professional tasks, and scientific research. GPT-6.1 Sol offers what Google Cloud calls near-Astra intelligence at lower cost. GPT-6 Luna is the fast, cost-efficient option for everyday work at scale. Google Cloud also says it’s now the only hyperscaler offering all three leading closed-source model families alongside more than 200 first-party, third-party, and open models.

Customers in sectors like banking and healthcare will be able to reach OpenAI’s models through Google Cloud’s access controls and data governance tooling to build agentic AI workflows. According to Google Cloud’s announcement, authored by Gemini product VP Michael Gerstenhaber, prompts, context, and outputs stay entirely within the customer’s data perimeter and are never used to train foundation models. Zero Data Retention options are coming soon. For CIOs whose compliance teams treat “sending data to OpenAI” as a four-letter word, that framing matters more than any benchmark score.

The Microsoft era is really over

Less than six months ago, Microsoft held the exclusive license to OpenAI’s intellectual property. That’s gone. Amazon started hosting OpenAI models in April, one day after the exclusivity expired, and Google’s arrival completes the trio: Azure, AWS, and now Google Cloud all sell the same frontier models.

AWS CEO Matt Garman told reporters at the time that customers had “long wanted to run OpenAI models where their applications and data already sat.” That line explains this deal better than any press release could. Enterprises don’t want to pick a cloud around a model. They want the model wherever their data already lives.

Google has been signaling this direction for a while. Thomas Kurian, the chief executive of Google Cloud, has told Reuters in a previous interview that Google’s cloud platform is an open platform and that it welcomes any model popular with developers. The company also struck a deal last year to give OpenAI additional computing capacity — back when the two were mostly described as rivals.

Why OpenAI and Google Cloud both want this

Follow the money and it makes sense for everyone except, arguably, Microsoft.

OpenAI is chasing enterprise revenue hard as it courts investors at a reported $1.4 trillion valuation ahead of a planned IPO. Direct API sales only go so far. Selling through hyperscaler marketplaces puts OpenAI’s models in front of procurement teams that buy software the same way they buy servers — through a cloud bill. According to data from corporate card issuer Ramp, most business spending on AI already flows to OpenAI and Anthropic, so OpenAI is essentially meeting that demand where it already sits.

Google Cloud, meanwhile, just posted the hottest growth numbers in the cloud business. Revenue growth accelerated to about 82% in the second quarter of 2026, up from 63% in the first quarter, driven by AI workloads and custom chips. Amazon’s cloud grew 37% and Microsoft’s 42% over the same stretch. Hosting OpenAI’s models fits Google’s pitch that its cloud is the neutral ground: the one place where you can run Gemini, GPT, and Anthropic models side by side. That pitch gets a lot stronger when one of those models is the most famous AI brand on the planet.

Why this matters

The real story here isn’t the partnership. It’s what the partnership confirms. Six months after exclusivity ended, the cloud wars over AI models have converged on the same answer: every hyperscaler sells everything.

That’s good for buyers. It means model choice is becoming a line item on a cloud bill, negotiated with volume discounts and committed-use pricing, instead of a strategic bet locked to one vendor. A team can prototype on GPT-Astra in Model Garden, switch to Gemini 4 Argon for a different workload, and never migrate off Google’s infrastructure.

But it also shows who holds the cards now. OpenAI has the models everyone wants, and every cloud wants them. When your product is the one customers demand by name, you don’t need exclusivity. You need distribution. And right now, OpenAI has all three clouds doing its distribution for it.

All three providers — Amazon, Microsoft, and Google — report quarterly results next week. Expect OpenAI’s name to come up on every call.

FAQ

When will OpenAI models be available on Google Cloud?

A private preview for select enterprise customers starts today; general availability is expected in the coming weeks. Models are sold through Google Cloud Marketplace and deployed via Model Garden.

Which OpenAI models are coming to Google Cloud?

GPT-Astra (flagship coding and research work), GPT-6.1 Sol (near-Astra intelligence at lower cost), and GPT-6 Luna (fast, cost-efficient everyday work), plus the Codex coding agent. Local Codex CLI and Desktop workflows can connect directly to Google Cloud’s platform for inference.

Can companies pay with existing Google Cloud commitments?

Yes — customers can apply their existing cloud spending commitments to OpenAI model purchases, according to the companies’ announcement.

What about regulated industries?

Banks, healthcare companies, and others in regulated sectors will get access through Google Cloud’s access controls and data governance tools to build agentic AI workflows.

Sources: Reuters (Krystal Hu), Unite.AI, newsbytesapp.com, LavX News