Prefill vs decode is why two people can run the same model on the same machine and disagree about whether it is fast. One waits ten seconds for the first word, the other watches text stream instantly, ...