Google Cloud is advancing its Enterprise AI strategy by announcing the availability of OpenAI models through Model Garden on the Gemini Enterprise platform. This move offers businesses access to leading AI models from top developers within a unified system, emphasizing strengths in security, data governance, and cost management to accelerate the transition of AI from experimentation to practical organizational use.
This initiative makes Google Cloud the sole hyperscale cloud provider supporting closed-source large language models (LLMs) from all three leading developers by adding OpenAI to its ecosystem, which already includes Google’s own models and Anthropic’s Claude model family. Customers can now select from over 200 AI models, encompassing both commercial and open-source options.
Google Cloud states that integrating OpenAI models via Model Garden allows businesses to flexibly choose or switch models based on task requirements, considering factors such as capability, processing speed, and cost. This supports a range of applications from software development to creating AI Agents capable of executing multi-step tasks autonomously.
Beyond accessing models through Google Cloud’s API, customers can purchase OpenAI models via Google Cloud Marketplace, consolidating expenses into a single Google Cloud bill. These costs can also be combined with long-term usage contracts and Flexible Savings Plans (FSP) benefits.
This move reflects Google Cloud’s strategy to be more than a provider of its own AI models; it aims to serve as a central infrastructure enabling organizations to access and manage models from multiple providers on a single platform without vendor lock-in or costly architectural changes when switching models. This reduces complexity in procurement, system development, and AI budget control.
A key challenge in the Enterprise AI market is deploying AI widely without increasing risks related to data security and compliance with organizational regulations.
Google Cloud affirms that customer prompts, context, and outputs remain within organization-specific data boundaries and will not be used to train base models. They also plan to introduce a Zero Data Retention (ZDR) option that will not store data in the system moving forward.
Meanwhile, OpenAI models operate on Google Cloud infrastructure with centralized identity verification, access controls, and usage audit logs, enabling organizations to monitor AI Agent activities throughout processes. Support for Confidential Computing technology is also being prepared.
These features target large enterprises that need to apply AI to critical data and workflows while maintaining corporate governance and security standards.
Greg Brockman, OpenAI’s President and co-founder, stated that Google Cloud is a trusted platform for developers and enterprises to build AI Agents, develop software, and deploy AI solutions. This partnership allows customers to access OpenAI’s advanced models, including Codex, through an established platform.
Thomas Kurian, CEO of Google Cloud, noted that the goal is to offer diverse AI models so organizations can select tools best suited to specific tasks, enhancing output quality while controlling costs.
The GPT 6.1 Sol and GPT 6.0 Luna models are currently being made available to select enterprise customers for private preview testing via Model Garden. Google Cloud customers can access these models through APIs to develop their own solutions and integrate Codex CLI and Codex Desktop with the platform for AI model processing. Broader availability of OpenAI models through Gemini for Business and Google Antigravity will follow in due course.
Read technology and digital transformation innovation news with ThairathMoney to help you "Improve your finances and life." https://www.thairath.co.th/money/tech_innovation
Follow the Facebook page: Thairath Money at this link - https://www.facebook.com/ThairathMoney