If you don't have a software for it, you need to calculate the full partition function manually, i.e. the translational and rotational (at least in gas phase) as well as the vibrational and electronic (all phases) partition functions. All of their calculation or at least the standard approximations can be found in any statistical mechanics book or also on wikipedia, e.g. https://en.wikipedia.org/wiki/Rotational_partition_function

Note that the electronic partition function is often simply approximated by the electronic ground state degeneracy.

Then you can calculate the Helmholtz Free Energy as described here:

https://en.wikipedia.org/wiki/Partition_function_(statistical_mechanics)

and make the leap to the Gibbs Free Energy via Legendre transformation.

Just calculate interaction, solvation energies and entropy as well.

delE = non bonding + coulombic energies

delH = delE + delsolv

delG = delH - TdelS

If the free energy is an attribute of an ensemble, isn't the entropy included implicitly in the interaction energies?

Martin Klvana : in principle, you might say that this implicit relation is exactly what statistical mechanics expresses. The entire canonical ensemble is derived by maximizing the entropy with the constraint of energy conservation.

Jürgen Weippert : What if the arrangement is part of solids, like unit cell ? I have to consider vibrational and electronic partition function.

In that case how to calculate the electronic partition function ?

If it's a solid, you need to go via the density of states, see e.g. p. 15 on:

https://www.tkm.kit.edu/downloads/TKM_I_2011_lecture_notes.pdf

However, if you do not have a software, you would need an analytic expression for the density of states and unless you want to go ultrarough and use Einstein or Debye I am not aware of good concepts there.