[hnidiots3]

Codex isn’t even a good model.

[k4rli]

It's been praised here. However in my experience Sonnet4.5 has produced as good if not better results.

[conception]

What I discovered is different models are simply just better at different things and it’s very hard to predict which one will be better than the others at a certain task. Thus no one has similar experiences as no one is doing the same things the same way ever.

[7thpower]

Based on what? It’s been great for me.