OpenAI Models Land on Google Cloud, Now Available Across All Three Major U.S. Cloud Platforms
nashnova research
OpenAI struck a deal with Google Cloud, making its models available on all three major U.S. cloud platforms — AWS, Azure, and Google Cloud; beyond model power, enterprise data and governance capabilities may be the real moat.
What exactly is shipping?
Google Cloud customers can buy OpenAI's frontier models and the Codex coding tool through Model Garden and Marketplace, using existing cloud-spend commitments to pay.
OpenAI models will also plug into Google's Antigravity dev environment and Gemini Agent, expected to go live in the coming weeks.
This means → enterprise customers can run OpenAI models on Google Cloud without moving data or switching platforms.
How did OpenAI reach all three clouds?
Earlier this year Microsoft released its exclusive license on OpenAI's intellectual property; AWS announced OpenAI model hosting the very next day.
Amazon has since expanded the OpenAI models and services on its Bedrock platform, but neither company has disclosed the additional revenue generated.
In plain terms = Microsoft's "unbundling" was the starting gun. AWS jumped in immediately, Google Cloud fills the last slot — OpenAI went from exclusive to full coverage in roughly six months.
Why would Google sell a competitor's models?
Analyst Mike Leone noted that many Google Cloud customers want to use OpenAI models without migrating their data — serving that need keeps them on Google's platform.
He read the deal as "a sign of confidence" — Gemini now has to win customers with OpenAI sitting right next to it.
The key: whichever AI model a customer picks, Google still collects fees for cloud infrastructure, data services, and security — Google gets paid either way.
How fast is Google Cloud itself growing?
Per CNBC, Google Cloud's Q2 annualized revenue growth hit roughly 82%, up from 63% in Q1, partly driven by strong demand for AI models and custom chips.
AWS and Azure posted Q2 cloud-revenue growth of about 37% and 42%, respectively.
This means → Google Cloud is the fastest-growing of the three; bringing in OpenAI models looks more like a leader doubling down than a defensive move.
When every cloud has top-tier models, where is the real moat?
OpenAI's core logic: enterprise AI competition hinges not just on model capability but on whether models can be deployed where enterprise data and applications already live.
Yet Leone also flagged the limit: once every cloud carries top-tier models, the edge goes to whoever manages enterprise data governance and cost control best.
In plain terms = models are becoming table stakes. The cloud provider's existing enterprise relationships and infrastructure may matter more than the models themselves.
What does this mean for OpenAI?
OpenAI is currently seeking funding at a valuation of roughly $1.4 trillion and preparing for an IPO.
Full coverage across three clouds is a distribution milestone, but whether it converts into real enterprise-customer growth is the true test of commercial value.
This reflects a broader shift: OpenAI is moving from "technology leader" to "enterprise platform supplier" — the next chapter of its valuation story needs revenue to back it up.
市场有风险，内容仅供研究参考，不构成投资建议。