The partnership will let enterprises buy OpenAI models through Google Cloud Marketplace, while Microsoft remains OpenAI’s primary cloud partner.
OpenAI is partnering with Google Cloud to make its frontier AI models and Codex coding agent available through Google Cloud’s Model Garden, giving enterprises another route to deploy OpenAI technology within their existing cloud environments.
The integration will allow Google Cloud customers to purchase OpenAI models through Google Cloud Marketplace and use them to build AI agents, support software development and run enterprise AI workloads. Organizations with existing Google Cloud Platform (GCP) spending commitments will be able to apply those commitments toward purchases of OpenAI models.
General availability is planned for the coming weeks, OpenAI said in a statement. However, Microsoft remains OpenAI’s primary cloud partner, the statement added.
The agreement expands OpenAI’s distribution through a cloud provider that already serves enterprise developers and data science teams. By making its models available through Google Cloud’s existing marketplace and development environment, OpenAI is seeking to reduce the friction involved in evaluating and adopting its technology, while giving customers another option for accessing frontier models.
Enterprise governance and model choice
The integration will combine OpenAI models with Google Cloud’s access controls, centralized data governance and computing infrastructure. OpenAI said these capabilities are intended to support demanding enterprise workloads, including deployments in regulated industries where organizations need to control access to sensitive data and how it is used.
The announcement, however, does not specify data retention terms, model training permissions or the precise security controls that will apply to the integration. Enterprises will need to examine those details before using the service with sensitive workloads.
The partnership also extends model choice within Google’s AI development ecosystem. OpenAI models are expected to become available in Google Antigravity and Gemini Agent, giving enterprise customers additional options when selecting models for coding and other tasks.
The move reflects a broader enterprise requirement: access to multiple AI models without having to build separate procurement and deployment processes for each provider. Google Cloud CEO Thomas Kurian said the company wants customers to select models based on the requirements of individual tasks, with the aim of improving output quality and controlling costs.
For customers, the practical value will depend on how the models are priced, how they integrate with existing development tools and whether enterprises can apply their current cloud commitments to the workloads they intend to run.
Codex targets software development workflows
The inclusion of Codex puts software engineering among the partnership’s key use cases. OpenAI said the integration will bring its agentic workflows into Google Cloud, enabling enterprises to use AI in software development and other multistep tasks.
Unlike conventional AI assistants that primarily respond to individual prompts, agentic systems can carry out sequences of actions toward a defined objective. In software development, these workflows can involve interacting with code, development tools and related resources, making access permissions and oversight important considerations.
OpenAI has not specified which Codex capabilities will be available through Model Garden at launch or whether the offering will differ from other Codex deployment options.
Google Cloud’s access controls and governance features are intended to provide an enterprise foundation for these workloads. Those features do not eliminate the need for organizations to set appropriate permissions, monitor agent actions and assess the risks of connecting AI systems to code repositories and production environments.
The partnership also gives Google Cloud another model provider to offer customers as competition intensifies among cloud platforms to support enterprise AI development. For OpenAI, distribution through Google Cloud could make its models more accessible to organizations already operating on the platform.
OpenAI said the arrangement is intended to give enterprises a path from model evaluation to wider adoption. The companies have not announced a firm general availability date. Customers can request access through OpenAI’s Google Cloud partnership page, the statement added.
