I am a paid Google One AI Pro subscriber using Google AI Studio Playground workspace only.
Google recently removed the Gemini 2.5 Pro model from all subscriptions, forcing my active threads onto Gemini 3.1 Pro. The new model constantly hits a 2-minute thinking loop and throws an ‘Internal error has occurred’ server timeout in all existing threads. Even branching from my saved checkpoint doesn’t change anything. I am paying for a 1-million+ token context feature that is currently completely broken due to a forced backend migration.
I want my case to be escalated to the AI Studio infrastructure engineering team to restore thread migration stability.
Also this proves the weakness of Gemini 3.1 pro model in parsing and handling the long context, because Gemini 2.5 pro worked fine even in 900k tokens thread.
As I pay for 2.5 pro model and was forced to use 3.1 pro model without my own intention, I want to know when Google restore this model in all subscriptions.