Google Cloud adds OpenAI's GPT-6 models to its model garden

A team running on Google Cloud can now call an OpenAI model from the same console it uses for Gemini. Google Cloud said in a blog post that it is adding three OpenAI models to its model garden: GPT-6 Astra for demanding coding and research work, GPT-6.1 Sol as a lower-cost near-Astra option, and GPT-6 Luna for everyday work at scale. The company said it is the only hyperscale cloud provider now offering access to Gemini, Claude and GPT alongside more than 200 other first-party, third-party and open models.

What a developer gets

The three OpenAI models are rolling out in private preview to select enterprise customers in the model garden, with wider availability later. Developers can also point their local Codex CLI and desktop workflows directly at Google Cloud's Gemini platform for inference, so the same workflows they already use with OpenAI's API now resolve against Google Cloud infrastructure. Google Cloud said more will come, including OpenAI models inside its Gemini for business product and Google Antigravity.

How it is billed

OpenAI models are sold through Google Cloud Marketplace, which Google Cloud said means usage shows on a single Google Cloud invoice, draws down existing cloud commitments, and counts toward committed-use discounts. The practical effect is that a team already locked into Google Cloud spend can try the GPT models without a second vendor contract or a separate credit card.

What the pitch is

Google Cloud framed the launch as "choice" rather than a model race. Its argument is that modern enterprise AI needs the right model for each workload, not one model for every task, and that this is cheaper to arrange when every model sits behind the same APIs, the same identity, and the same audit trail. The post called out strict data perimeters, isolated access boundaries on dedicated infrastructure, and unified logging across every agent and model interaction. Yashodha Bhavnani, VP of AI Products at Box, was quoted saying this would let Box customers reach OpenAI's frontier models alongside Google Cloud's security and compliance controls.

For a team already deciding whether to port its agent from one lab to another, the move changes what a decision looks like. Picking a model used to mean picking a cloud too, because the cheapest path to a given model's API was its maker's own cloud, and switching meant redoing the authentication, the billing setup and the compliance review. If a GPT model and a Gemini model answer to the same identity, the same billing, and the same audit log, picking is now a per-workload choice rather than a company-wide commitment. The hyperscaler that offers the broadest set of models can keep the account for the longer term.

Source

Primary source: Google Cloud brings OpenAI models to customers, Google Cloud Blog.