For short circuit calculations, an infinite bus is going to be anything greater than what the transformer will let through it.
In this case, 5100A at 12.5KV equates to 110.4MVAsc
Without any impedance between the 12.5KV and 480V system, it equates to 132.8kAsc
Typical transformer impedance for a voltage less than 25kV primary and 480V secondary is 5.75%. So, 480V, x 3200A x 1.732 = 2.66MVA. Divide it by the impedance and you get 46.3MVAsc which equates to 55.6kA @ 480V, assuming infinite bus.
Even if you increase your service to 5000A at 480V, the maximum possible sc current on the 480V side (assuming 5.75% impedance) is going to be around 87kAsc. You could order a transformer with a 6.75% impedance and still get the SC current under 65kA.
The transformer is going to limit the fault current to it's own MVAsc capability on a system that assumes an infinite bus.
Actually, I have to retract the infinite bus statement, as further number crunching determined in using 3200A service, it is not. 110.4MVAsc combined with 46.3MVAsc equates to only 32.6MVAsc on the 480V side which is around 39.2kA. However, if the service is around 800A or less, it would appear to be an infinite bus.
As far as the OP, they did not provide enough information to model in SKM. You need either system impedance, i.e. R and X values, or the X/R ratio in the program.