I could always do both. I have the internal voice, that talks and thinks and analyses things all the time. But I can also willingly turn it off, pause it and think purely visually, which is much faster than narrating (I can do it in two languages). Thinking visually, like playing silent video or manipulating objects in your mind is useful, but in the end the inner voice always continues with some conclusion. The best example I can give is playing chess. I can play blind chess (no board or physical chess set needed). It is much faster to think visually. Move pieces in your head, imagine positions well in advance, take things back and explore different variants, than it is to narrate: if I play kings pawn forward two squares my opponent will play bishop to e2. Narrating is limiting in this case, and would slow things to a crawl. But when a particularly good movie is found through visual only thinking the voice still kicks in with “Aha that is a good move”. It’s inescapable.