# The last couple of hours, the gpt 3.5 turbo model does not respond in 1-3 seconds, but in 20-60 seconds. I use the python module, I haven't changed anything in the last couple of days