Model assortment while in the cell app is now broken. It doesn't matter which model you decide on, it gives you o1! This can be a challenge simply because o1 is far even worse than 4o at context, is slower, and fails to employ Memory, entirely breaking the Initiatives function. I've tried