@amerpie Ah, thanks for the heads up. If you change the model variable to “gpt-4o” it should continue to work. I’ll get an update out soon.