1) If you are looking for a face swap for single person, you might want to bypass the middle section of the workflow and enable the top section as you can see there is only one person inside.
2) For our sample 4 secs video in that face swap workflow, it took about 10 to 15mins to finish on a Large Machine, thus for your 12 secs video, it needs about 1 hour of Large Machine time. Extra Large will shorten the time about 20 to 30% comparing to Large last time I tested it.
3) Looking for more detailed explanation on FaceSwap? Check out our tutorials here, https://learn.runcomfy.com/guide-to-using-comfyui-reactor-workflow-for-video.