Google Cloud said on Oct. 21 that OpenAI’s GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna are rolling out in private preview to select enterprise customers in its Model Garden. Customers can also connect local Codex CLI and Desktop workflows to the Gemini platform for inference starting today, according to the Google Cloud announcement.
The commercial terms are the part a procurement team will read first. Google says OpenAI models are transacted directly through Google Cloud Marketplace, so usage appears on a single Google Cloud invoice, draws down existing cloud commitments and counts toward Flexible Savings Plans. It also says prompts, context and outputs stay inside the customer’s data perimeter and are never used to train foundation models. Zero Data Retention options and Confidential Compute are both listed as coming soon. The post gives no regions, no prices and no dates for general availability.
Michael Gerstenhaber, Google Cloud’s VP of product management for Gemini, wrote that Google Cloud is “now the only hyperscale cloud provider offering access to all three leading closed source LLMs.”
Amazon’s version of this offer is further along. In a June 1 post, AWS said GPT-5.5, GPT-5.4 and Codex were generally available on Amazon Bedrock, that “Pricing matches OpenAI first-party rates,” and that Codex usage counts toward existing AWS commitments. That post covers earlier models than the ones Google lists today, so a like-for-like price comparison has to wait. What the two posts do show is a generally available, priced offer on one cloud and a private preview with no published price on the other.
A buyer holding a Google Cloud commitment should find out whether OpenAI usage burns it down at first-party rates once pricing is published. Ask for the Marketplace terms in writing before planning workloads around these models. Our coverage of Gemini 4 Argon introductory pricing noted the discount ends Nov. 30, and the Gemini Agent’s spend caps and agent identities cover cost controls for agent workloads on the same platform.
Source: Google Cloud
