Comment by aleqs

11 hours ago

Opposite of my experience - openai models, and codex are clearly the slowest (out of 5 different models/harnesses I use on a regular basis).

0 comments