Finally, I no longer have to be forcibly downgraded to Lite; the automatic memory model really resonates with me.

Google Gemini announces six quota optimization measures to alleviate quota consumption: 1 Free Flash-Lite prompts; 2 Fix multimodal video deduction and double the Omni video quota; 3 Set deduction caps for Pro requests involving large files/long prompts; 4 Multimodal errors and failed requests are exempt from deduction; 5 Heavy tasks will provide transparent usage dashboards and overdraft alerts; 6 Automatically remember the user's selected model, and unless the quota triggers downgrade, subsequent switches to lightweight versions are not enforced.