Google Cloud is adding OpenAI’s advanced models to its enterprise AI platform, allowing companies to buy and operate technology from one of Google’s principal AI rivals through existing cloud contracts.
GPT-Astra, GPT-6.1 Sol and GPT-6 Luna are rolling out in private preview to selected customers through Google Cloud’s Model Garden, Google announced today. Customers will transact through Google Cloud Marketplace, place the charges on a single cloud invoice and apply existing spending commitments and Flexible Savings Plans to the purchases.
The arrangement lowers a procurement barrier for companies that already run their data and applications on Google Cloud. It also lets developers select and switch among OpenAI, Google, Anthropic and other models through a common platform rather than negotiate separate vendor contracts and build different governance systems for each provider.
Google said prompts, context and outputs from OpenAI workloads will remain within customers’ designated data perimeters and will not be used to train foundation models. The models will run on dedicated Google Cloud infrastructure with identity controls, audit logging and VPC Service Controls. Zero-data-retention options and Confidential Compute are planned but are not yet available.
Customers can begin using the models through Google Cloud APIs and connect local Codex command-line and desktop workflows to the platform. Google said OpenAI models will later become available through Gemini for business and Google Antigravity.
The partnership expands OpenAI’s distribution beyond Microsoft and Amazon. OpenAI’s exclusive distribution arrangement with Microsoft expired in April, and the company subsequently broadened access through AWS.
Google described itself as the only hyperscale cloud provider offering all three leading closed-source model families alongside more than 200 other models. The companies did not disclose commercial terms, pricing, participating customers or deployment results.