- Oracle is adding Gemini to OCI, starting with Gemini 2.5
- Further models, including specialized ones like MedLM, will also be added
- OCI recently signed a similar deal to add xAI’s Grok, too
Oracle and Google Cloud have announced a deal to make Google’s Gemini models available via Oracle Cloud Infrastructure’s (OCI) Generative AI service.
Oracle confirmed the service would kick off offering Gemini 2.5, but the company would continue to add further models for text, image, video, speech and music generation as well as specialized models like MedLM for healthcare.
The company explained customers would now be able to use Gemini models – among others – to help them build their own AI agents for coding, automation, research and more.
Oracle Cloud now offers Gemini models
“Oracle plans to make Google’s entire range of Gemini models available via OCI Generative AI service through new integrations with Vertex AI,” Oracle added.
Although the models will continue to run on Google’s servers, they can be accessed through Oracle’s cloud platform with customers paying with Oracle Universal Credits, but the benefits of this partnership will be twofold.
For Oracle, it means the company can expand its offering of models, including third-party options, rather than pushing its own in-house models, making it more attractive for broader audiences. With the deal, Google can also expand its enterprise reach, getting Gemini models into more hands.
“Oracle has been intentional in offering model choice curated for the enterprise, spanning open and proprietary models,” OCI President Clay Magouyrk explained.
The news comes around two months after Oracle struck a similar deal to make xAI’s Grok models available to customers, and while many hyperscale cloud providers are battling it out to secure exclusive or preferential AI model partnerships (like Microsoft with OpenAI and AWS with Anthropic), Oracle’s multi-vendor strategy could give it a more stable position in the future.
