Randomization across 12 versions of 2 tasks

My study has 2 tasks, one that has 12 versions and hte other that has 6. Subjects must participate in both tasks over 4 sessions. I would like to create individual links per version so that I can control the randomization (we have our own existing randomization per subject), but that would involve creating 18 different online experiments and each subject will complete 4 within a study.

E.g For a single subject their participation involves:

Session 1 = WPA (Version 1 ) + OSPAN (Version 3)
Session 2 = WPA (Version 1 ) + OSPAN (Version 4)
Session 3 = WPA (Version 6 ) + OSPAN (Version 2)
Session 4 = WPA (Version 6 ) + OSPAN (Version 9)

My questions are:

  1. Do others recommend I create separate experiments for each version?
  2. Is the cost still the same .2 for all 4 sessions or would I be charged per session?
  3. Suggestions on how to make this simple and easy for subjects?

Many thanks,
Kate