Google Cloud adds OpenAI models alongside Gemini and Claude

Google Cloud is adding OpenAI models to a lineup that already includes Gemini and Claude, according to The Decoder. The change matters for enterprise AI buyers because Google Cloud says it is now the only major cloud provider offering all three model families on one platform.

For teams that have been weighing ChatGPT, Google Gemini and Claude separately, the announcement points to a more consolidated way to test and procure leading AI models. It also puts Google in the unusual position of selling access to models that compete directly with its own Gemini work.

What Google Cloud is adding

According to the report, three OpenAI models are available through Google Cloud as part of the launch:

• GPT-6 Astra, described as intended for demanding work such as coding and research.
• GPT-6.1 Sol, described as a lower-cost alternative with similar performance.
• GPT-6 Luna, described as a fast, cost-efficient model for everyday tasks.

The models are launching in private preview for selected enterprise customers, with sign-ups open now. The Decoder reports that the models will also appear in Gemini for Business and Google Antigravity.

Google Cloud says charges for the OpenAI models will run through a single Google Cloud bill and count toward existing contracts. The company also says customer data will not be used for training.

For enterprises, those billing and contract details may be as important as the model names. AI adoption often stalls when teams have to route each provider through separate procurement, legal and security reviews. A single cloud relationship can make experiments easier to approve, even if the technical evaluation still has to happen model by model.

Why this matters for enterprise buyers

The practical shift is model choice inside one procurement path. A product team could evaluate Gemini, Claude and OpenAI models without necessarily opening three separate vendor relationships, based on the information Google Cloud has provided.

That does not remove the hard work of choosing the right system. Businesses still need to test outputs, latency, reliability, governance fit and cost against their own workloads. But the platform approach may make it easier for centralized AI teams to offer approved options to departments that want different models for different jobs.

This is especially relevant for companies building internal AI assistants, coding workflows, research tools or marketing operations. Some teams may prefer one model family for writing, another for analysis and another for software tasks. The announcement suggests Google Cloud wants to support that mixed-model reality rather than forcing customers into a single default.

Readers comparing the broader landscape can also see our ChatGPT vs Claude vs Google Gemini buying guide for a practical framework, though Google Cloud’s new preview is specifically about enterprise access through its platform.

Google’s balancing act

The move also highlights a tension inside Google, according to The Decoder: compute capacity used to serve cloud customers can also support companies competing with Google’s own Gemini team. In this case, Google Cloud is broadening customer choice while hosting models from OpenAI and Anthropic’s Claude line beside Gemini.

The report notes that this kind of resource conflict has been a recurring issue at Google and says it reportedly played a role in researcher Noam Shazeer returning to OpenAI. The source identifies Shazeer as a two-billion-dollar researcher, but does not provide further detail in the article text.

For Google Cloud, the strategic logic is straightforward: enterprise customers increasingly want access to multiple AI systems, and cloud platforms want to be the place where that access is managed. For Google’s AI product teams, the optics are more complicated. The cloud business benefits when customers can buy whatever model they want, even when that includes Gemini’s rivals.

That tension is not unique to Google, but it is especially visible here because Gemini is both a product Google wants customers to adopt and one option among several now being offered through the same cloud platform. Google recently expanded its own enterprise Gemini work as well, including an enterprise Gemini agent for work across apps.

What to watch in the preview

The current rollout is not a broad public launch. The Decoder says access begins as a private preview for select enterprise customers, so smaller businesses and independent creators should not assume they can use these OpenAI models through Google Cloud immediately.

The main things to watch are whether Google expands availability beyond the preview group, how the models are presented inside Gemini for Business and Google Antigravity, and how enterprises respond to buying competing model families through one bill.

For small businesses, the near-term impact is indirect. This is primarily an enterprise procurement story, not a new consumer chatbot release. But the direction is important: AI buyers are pushing for flexibility, and cloud platforms are responding by packaging rival model families side by side.

If that pattern continues, the winning workflow may be less about picking one model forever and more about choosing a platform that lets teams switch models as tasks change.