OpenAI’s decision to offer GPT models through Google Cloud is best read as a distribution move made possible by its release from exclusive cloud licensing: developers will be able to access the company’s artificial intelligence models through Google’s cloud in the coming weeks.
OpenAI will place its GPT products in Google Cloud’s Model Garden, Marketplace and Antigravity coding tool. Customers will be able to use cloud spending commitments they already hold to pay for them.
The arrangement gives Google Cloud a route to sell OpenAI’s models beside its own Gemini offerings. For OpenAI, it means its models are offered across all three leading cloud infrastructure providers.
The shift follows a renegotiation with Microsoft. Less than six months ago, OpenAI said Microsoft would no longer hold the exclusive licence to its intellectual property. Under the revised agreement, OpenAI gained the right to cap its revenue-sharing obligations and offer services through rival cloud platforms.
Amazon moved quickly after that change. The day after Microsoft’s exclusive licence ended, Amazon announced plans to host OpenAI models on AWS Bedrock. OpenAI then announced a limited-preview release of its newest models and its Codex coding agent on the Amazon Web Services platform in April, one day after the renegotiated Microsoft deal changed its cloud options.
Before that April announcement, AWS developers could access only OpenAI’s open-weight models, which arrived on the platform in August 2025. Matt Garman, Amazon Web Services’ chief executive, said at the launch event that customers had “asked for the offering for a long time”.
The cloud platforms are competing for AI workloads while seeking to meet demand for the models that businesses are buying. Most corporate spending on AI goes to OpenAI and Anthropic, according to data from the corporate card issuer Ramp.
Google Cloud’s annualised revenue growth in the segment accelerated to about 82% in the second quarter, from 63% in the first quarter. Amazon reported second-quarter cloud growth of 37%, while Microsoft reported 42%.
Google released its next-generation Gemini 4 Argon model in September. It will now sell that offering alongside models from OpenAI, a company founded in 2015 partly as a counterweight to the concentration of AI research at companies including Google.
The companies have worked together before. Last year, OpenAI said it was using Google’s cloud to run ChatGPT and its application programming interface.
Amazon has also deepened its commercial relationship with OpenAI. It pledged $50 billion to OpenAI roughly three months after disclosing a $38 billion commitment between the two companies in November 2025.
For OpenAI, access through Google Cloud adds another sales and billing channel as it courts investors at a $1.4 trillion valuation and targets an initial public offering. The company has yet to specify which GPT models Google Cloud will offer, their pricing, the division of revenue between the companies, or a timetable for general availability beyond the coming weeks.
Comments (0)