Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseInsight · On RadarOpenAI Blog→ original

OpenAI Models and Codex Are Coming to Google Cloud — You Can Pay From Your Cloud Budget

OpenAI and Google Cloud have agreed that OpenAI's flagship models and Codex, its AI coding agent, will appear in Model Garden, Google Cloud's model catalog. Customers can pay through Google Cloud Marketplace, drawing from a cloud spending commitment they have already agreed to (GCP, Google Cloud Platform). General availability is promised within the coming weeks, so teams on GCP should start preparing tasks for comparing models.

AI-processed from OpenAI Blog; edited by Hamidun News
OpenAI Models and Codex Are Coming to Google Cloud — You Can Pay From Your Cloud Budget
Source: OpenAI Blog. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceOpenAI Blog ↗
Editor-in-Chief
Key metric #1Payment goes through Google Cloud Marketplace against GCP spending commitments
Key metric #2Later, the models will appear in Google Antigravity and Gemini Agent
Key metric #3Microsoft remains OpenAI's primary cloud partner

Imagine a company that has spent years building data and AI pipelines on Google Cloud. The business wants OpenAI models connected, and the familiar bureaucracy begins: a separate contract, a separate invoice, a new security review, and arguments over where the data will go. Now OpenAI and Google Cloud have agreed to remove some of that friction. OpenAI's flagship models and Codex (OpenAI's agent that writes and edits code) will appear in Model Garden, the model catalog inside Google Cloud. Customers will be able to pay for them from a cloud budget they have already agreed to, that is, from spending commitments on GCP (Google Cloud Platform, the name of Google's cloud platform).

What was actually announced

OpenAI and Google Cloud are bringing OpenAI's flagship models and Codex agentic workflows to Model Garden. Customers will get native access to them directly in the environment where developers and data scientists already work. General availability is planned for the coming weeks, and customers can request access now.

A separate part of the announcement concerns money. The models can be purchased through Google Cloud Marketplace, Google's cloud services store. For companies that have already committed to spending a set amount on GCP, these purchases will be charged against those commitments. For a finance director, this is a strong argument: there is no need to open a separate budget line, and cloud volume already paid for under contract does not go to waste.

In addition, OpenAI models will soon appear in Google Antigravity and Gemini Agent. Inside Google's own tools, users will be able to choose a model for the task rather than using only Gemini.

Why it matters for regulated industries

The announcement puts considerable emphasis on data. Banks, healthcare providers, and government agencies cannot simply send confidential documents to an external API. What matters to them is who has access to sensitive data and how it is used. According to the companies, OpenAI models will work together with the access controls built into Google Cloud, centralized data governance, and high-performance infrastructure.

The message contains no technical details: no regions, model versions, prices, or data retention terms are named. The phrase about data governance should therefore be read as a stated intention, not a finished guarantee. Alongside this, OpenAI's page mentions its own separate steps: zero data retention for flagship models, and access to the models and Codex through cloud commitments with Oracle. It is clear that the company is steadily embedding itself in the cloud contracts of large customers, rather than only selling its API directly.

What it means for developers and product managers

First, teams on GCP get a simple way to compare models. If you use Gemini but suspect an OpenAI model would be a better fit for code generation or complex agent chains, you can test that without a new procurement process or a separate integration. Thomas Kurian, head of Google Cloud, states the goal directly: give customers a choice of model for the specific job, so they get the best quality and lower costs.

Second, Codex becomes more accessible to enterprise teams that build their development processes on Google Cloud. Greg Brockman, OpenAI's president and co-founder, says the company wants to offer its models and Codex where customers are already building their systems.

Third, the calculation changes for founders and procurement teams. Choosing among several models is no longer only a technical question: it can also be shaped at the level of cloud contracts.

Limitations and context

The scale should not be overstated. Access is not yet open to everyone, and there is neither a price list nor a list of specific models. OpenAI notes that Microsoft remains its primary cloud partner, so this is an expansion of sales channels, not a change of main ally.

The deal is also telling for Google. The cloud business competes for enterprise AI budgets, and its willingness to host the models of Gemini's main rival shows that customers want to choose rather than be tied to a single lab.

What to do now: if you work on GCP, request access, check the data and pricing terms once they are published, and assemble a set of your own tasks for comparing models in advance. Choose the model that handles your data better and costs less, not the one with the louder brand.

Story

Key Questions Answered

What was actually announced?

OpenAI and Google Cloud are bringing OpenAI's flagship models and Codex agentic workflows to Model Garden. Customers will get native access to them directly in the environment where developers and data scientists already work. General availability is planned for the coming weeks, and customers can request access now. A separate part of the announcement concerns money. The models can be purchased through Google Cloud Marketplace, Google's cloud services store.

Why it matters for regulated industries?

The announcement puts considerable emphasis on data. Banks, healthcare providers, and government agencies cannot simply send confidential documents to an external API. What matters to them is who has access to sensitive data and how it is used. According to the companies, OpenAI models will work together with the access controls built into Google Cloud, centralized data governance, and high-performance infrastructure.

What it means for developers and product managers?

First, teams on GCP get a simple way to compare models. If you use Gemini but suspect an OpenAI model would be a better fit for code generation or complex agent chains, you can test that without a new procurement process or a separate integration. Thomas Kurian, head of Google Cloud, states the goal directly: give customers a choice of model for the specific job, so they get the best quality and lower costs.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Hands-On AI Engineering · Hamidun Academy

Build & Ship Production AI Agents Without Boilerplate Code

Start your free 72-hour trial at Hamidun Academy (academy.hamidun.com) — interactive lessons on Claude Code & Cursor, live Tech Radar, and production agent templates.

Start 72h Free Trial →Vibe-Coding Intensive
What do you think?
Loading comments…