Gemini 2 Flash has been announced Production-Ready but did not (VertexAI) offer quota limits. As far as I have been informed, the only way would be buying provisioned throughput (unreasonably expensive). Is there any possibility the model quota will work like the 1.5 versions?