Enterprises can use a powerful, near topline OpenAI LLM on their hardware totally privately and securely, without sending data to the cloud.

The Gemini 2.5 Deep Think released to users is not that same competition model, rather, a lower performing but apparently faster version.