AI Industry

OpenAI Models Are Coming to Google Cloud Model Garden

Oct 21, 2026·5 min read·By Rahish Saifi

A cloud data centre aisle with rows of black server racks, a central open rack of servers wired with orange and blue cables, and a closed laptop on a steel cart under amber cable trays.

OpenAI and Google Cloud said on October 21, 2026 they are partnering to bring OpenAI's frontier models and Codex agentic workflows into Google Cloud's Model Garden. The move will give enterprise customers native access to those models inside the Google Cloud environment they already use for data, security, and billing—while OpenAI continues to treat Microsoft as its primary cloud partner.

Key facts

Item	Detail
Announcement	October 21, 2026 — OpenAI blog post and company statements reported by CNBC and Reuters
What ships	OpenAI frontier models and Codex agentic workflows in Model Garden
How to buy	Purchase via Model Garden, transacted through Google Cloud Marketplace
Procurement	Enterprises with Google Cloud spend commitments can draw down commitments for these purchases
Governance	Pairing with Google Cloud access controls, centralized data governance, and high-performance infrastructure (per OpenAI)
Also coming	OpenAI models soon in Google Antigravity and the Gemini Agent
Availability	Private preview rolling out now to select enterprise customers (Google Cloud); general availability planned in the coming weeks (OpenAI)
Microsoft	OpenAI states Microsoft remains OpenAI's primary cloud partner

What OpenAI and Google announced

OpenAI frames the integration as a way to run complex agentic workflows at scale without leaving Google Cloud. Teams could modernize software development, build agents, and change how people work while staying on Google's ML and AI infrastructure—the same stack many data scientists and developers already use.

For regulated industries, OpenAI highlights control over who can access sensitive data and how it is used. The partnership pairs OpenAI's frontier models with Google Cloud's built-in access controls, centralized data governance, and infrastructure aimed at demanding enterprise workloads.

Google Cloud customers will buy OpenAI frontier models through Model Garden on Marketplace. Organizations with existing GCP spend commitments can apply those commitments toward the purchase, which matters for procurement teams that route AI spend through committed cloud budgets rather than new vendor contracts.

OpenAI also said its models will soon appear in Google Antigravity and the Gemini Agent, giving customers more model choice inside Google's own agent and coding surfaces—not only in standalone API calls.

“Google Cloud is a leading platform for developers and businesses building AI agents, writing software, and putting AI to work. 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.”

— Greg Brockman, President and co-founder, OpenAI

“We are focused on giving customers a broad selection of models, including those from OpenAI, so they can choose the right model for any given job, helping to deliver optimal quality and lower costs.”

— Thomas Kurian, Chief Executive Officer, Google Cloud

Why multi-cloud distribution matters

CNBC notes the deal extends OpenAI's push to reach enterprise developers beyond its consumer ChatGPT base. After OpenAI's exclusive distribution arrangement with Microsoft ended, Amazon began hosting OpenAI models in April, CNBC reports. With Google Cloud joining, OpenAI's models are on all three major U.S. cloud infrastructure providers.

Reuters describes hyperscalers as increasingly important commercial channels for frontier AI labs, with cloud providers typically taking a cut of sales through their platforms. For Google Cloud, adding OpenAI fits a strategy of offering a broader choice of third-party models rather than relying only on Google's own Gemini family—while the two companies still compete in core AI products and search.

CNBC also cites Google Cloud's recent momentum: annualized revenue growth in the cloud segment accelerated to about 82% in the second quarter from 63% in the first quarter, partly on demand for AI models and custom chips, with Amazon and Microsoft posting second-quarter cloud growth of 37% and 42%, respectively. All three report quarterly results the following week, per CNBC.

What it means for enterprises

If your organization is already standardized on Google Cloud, this partnership reduces friction: evaluation, procurement, and production can stay inside familiar Marketplace and commitment workflows instead of standing up a separate OpenAI billing relationship—subject to whatever terms Google and OpenAI publish at GA.

Security and compliance teams get a story that combines OpenAI's models with Google's access controls and data governance, which OpenAI explicitly calls out for regulated sectors. That does not replace your own data-handling review, but it aligns with how many enterprises already approve cloud AI services.

OpenAI's reminder that Microsoft remains its primary cloud partner signals that Azure-centric customers are not being deprioritized; Google is another distribution path, not a replacement headline partner.

What it means for developers

Developers on GCP can expect OpenAI frontier models and Codex agentic flows where they already discover models in Model Garden, with a path from evaluation to broader adoption, per OpenAI. CNBC says OpenAI's GPT products will also reach Google's Antigravity coding tool in the coming weeks.

Google Cloud says customers can already connect local Codex CLI and Desktop workflows to its Gemini platform for inference.

Model choice inside Gemini Agent means agent builders on Google's stack may pick OpenAI models for some tasks and Gemini for others—matching Kurian's pitch of the right model per job for quality and cost.

Google Cloud's own announcement names GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna, which are rolling out now in private preview to select enterprise customers in Model Garden, with more coming soon. OpenAI's post does not name specific models; Reuters cites GPT-6 Astra as an example.

What we don't know yet

Pricing — None of the cited announcements publish per-token or subscription pricing for Model Garden access.
Exact model list — Google Cloud lists GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna in private preview, “with more coming soon”; the full GA lineup is unconfirmed.
General availability date — OpenAI says GA is planned in the coming weeks without a calendar date. CNBC similarly points to Antigravity availability in the coming weeks.
Regional rollout and quotas — Not specified in the sources reviewed here.

Bottom line

OpenAI is meeting Google Cloud customers where they already build: Model Garden, Marketplace, spend commitments, and governance tooling—plus upcoming access in Antigravity and Gemini Agent. For enterprises, it is another signal that frontier models are sold through every major U.S. cloud, with procurement and compliance woven into the platform. Developers should watch for GA in the coming weeks and validate which models and prices actually land in their projects.