Nope. Think it has something to do with composer context getting too long cuz I can still make a new composer chat and talk to the AI.
But this is unfortunate because I can only talk to gemini-2.0-flash-thinking-exp-1-21 like 5 messages per composer chat and this model is very quick at thinking and outputting!
Also, please fix the formatting for gemini2.0flashthinkingexp1-21. It more times than not just writes code as a message in composer rather than within a code block that can be auto-applied to my codebase, so even when this api error doesn’t occur, I still can’t use the model since the code is unapplyable to my codebase.
This is most likely because you used a language other than English, such as Chinese. This is a problem with Gemini, and it has occurred before. However, it has reappeared now and has lasted for a very long time.