Basically, a reasonably good CPU and an extremely powerful GPU card to create those renders, like a RTX3080 of 3090 . You can make do or experiment with a lesser card, but not if you want to push out a long story. If you are really doing this for money, you invest in several computers or you connect to a render farm.
Still. I recently played a game that had 10 or 12 static locations. 5 characters with basic stances. But the dialogue made it all worth it.