That's one of the reasons I'm relying more on Gemini and Claude these days.

True but there is the way try to tun in on phone there it works smooth

LLMS struggle as the context window fills up. Open a new chat.

Not much. It's up to OpenAI to optimize in the future.