Comment by wahnfrieden

Codex’s model is much better at actually reading large volumes of code which improves its results compared with CC