On: Gemini 2.0: our new AI model for the agentic era

[petra]

But given inference time compute, to give a strong reply reasonably fast, you'll need a lot of compute, very rarely used.

Economically this fits the cloud much better.