OpenAI Brings GPT-6 Models and Codex to Google Cloud, Expanding Enterprise Reach Beyond Microsoft Azure

OpenAI is bringing its frontier models and Codex coding workflows to Google Cloud’s Model Garden, extending enterprise access through one of its strongest infrastructure rivals. The products will be offered through Google Cloud Marketplace, with availability beginning in the coming weeks, according to OpenAI and Google Cloud announcements.

The initial lineup includes GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna. Customers will be able to access OpenAI models alongside Google’s Gemini offerings within the same cloud environment, giving businesses more choice when building applications, agents and software-development tools. Techstrong, Unite and Ars Technica AI all described the move as a significant expansion of OpenAI’s enterprise distribution strategy.

Google Cloud will distribute OpenAI technology while competing with OpenAI in models, developer platforms and enterprise AI services. That arrangement puts GPT models inside a marketplace designed to keep cloud customers within Google’s infrastructure, even when they select a non-Google model for a particular workload.

The deal completes OpenAI’s presence across the 3 major U.S. hyperscalers. OpenAI models already serve customers through Microsoft Azure, while the company has also pursued access through Amazon Web Services. Finance.yahoo and Tech-insider framed the Google Cloud launch as a major step in making OpenAI’s products available across the cloud market rather than tying them to a single provider.

The marketplace launch follows a broader infrastructure relationship between OpenAI and Google Cloud. Google’s data centers will provide computing capacity for OpenAI’s training and deployment needs, adding another supplier as demand for model development and ChatGPT services continues to grow.

OpenAI’s expansion beyond Microsoft Azure reflects the scale and urgency of its computing requirements. CNBC reported in 2025 that OpenAI planned to draw on resources from Google, Microsoft, CoreWeave and Oracle. Reuters, cited in coverage from Finance.yahoo and Cloudwars, described the Google agreement as an unusual infrastructure partnership between companies competing for AI customers.

The Google Cloud arrangement reduces Microsoft’s position as OpenAI’s exclusive commercial gateway. OpenAI and Microsoft revised their partnership, opening room for OpenAI to sell its services through additional cloud providers, according to VentureBeat AI. Microsoft remains a major infrastructure partner and investor, but OpenAI now has more routes to distribute models and secure computing capacity.

That shift changes the balance between the companies. OpenAI gains negotiating power and broader access to enterprise buyers, while Microsoft faces competition from a cloud provider that also operates a rival model family. Revolgy described the agreement as a move away from OpenAI’s Azure-only posture, while Cloudwars connected it to rising demand and strain in the long-running Microsoft relationship.

Enterprise customers gain a single cloud purchasing and deployment path for both Google and OpenAI models. Access through Model Garden can simplify procurement, billing, security controls and integration for organizations already using Google Cloud, while Codex adds a software-development option to the model catalog.

The practical effect will depend on pricing, regional availability, performance and the technical controls attached to the preview. OpenAI’s announcement emphasizes frontier models and Codex workflows, while Google Cloud positions Model Garden as a place where customers can compare and deploy models from multiple providers. That creates room for companies to match models to specific tasks instead of standardizing every application on one vendor.

The launch puts model distribution at the center of cloud competition. Google Cloud can attract customers with a broader catalog, OpenAI can reach buyers that already operate on Google infrastructure, and enterprises can negotiate across more providers. The arrangement also gives Google a way to participate in OpenAI workloads even when Gemini isn't selected.

The next stage is availability and adoption. OpenAI and Google Cloud said the models will arrive in the coming weeks, while OpenAI is accepting requests for preview access. Early enterprise uptake will show whether customers treat the offering as a supplemental model option or shift meaningful workloads away from Azure and other platforms.