The Final Cut Pro alternative that finishes the chores for you
Credit where it is due: Final Cut Pro 11 added Transcribe to Captions, and it runs on the machine, downloading its language model once on first use. That is the same shape as Montaj's approach, and it means the usual on-device argument does not apply here.
What Final Cut still leaves to you is everything around the transcript. Silences are yours to find. Voice cleanup is yours to dial in. Word-by-word karaoke captions are a design job with keyframes. Montaj does all three by default, then gives you a real timeline for trims, reordering and B-roll layers with 100-step undo.