If you have relatively accurate masses in CAD it would be fairly trivial to someone familiar with forces and statics. Select all the components that are going to be rotating, put the system in the, “worst case,” position (COM horizontal from pivot point, or closest point to that) and determine the torque the system applies at the pivot point. Using that torque and pistons of various bores, calculate the minimum distance the piston needs to be from the pivot taking into account the angle at which the piston is applying the force.

Without accurate masses in CAD it just becomes a longer, more drawn out problem. You would need to determine the torque applied by each member and component when the system is in the worst case scenario (including rollers, motors, bearings, etc. depending on how accurate you want to be) to determine the torque on the pivot, and then follow the same process as above.

Arguably the most accurate method, assuming you are alright with not having everything figured out in CAD, is to measure the force it takes to lift/move the actual system from its worst case point (ideally tangentially to the point), calculate the torque that corresponds to the force at that distance, and solve for the minimum point of the cylinders again.

You can also simply overestimate the mass/torque of the system and size it that way which will likely be the quickest brute-force method. If you know your arm weighs roughly 10 lbs, and the COM can’t be further than 6" from the pivot, use something like 12 lbs and 7". Quick and dirty, but gets the job done.