Overview of Gaussian Theories

Gaussian-n theories (n=1,2,3 . . .) are a family of composite techniques
designed to reproduce the energies of molecular systems to within chemical accuracy.
The methods vary in computational effort as well as the accuracy obtained for a set
of experimentally known systems (test set). The main members of the family are:

• Gaussian 2 theories
  • G2 - the original
  • G2(MP2) - saving some time
  • G2(MP2,SVP) - saving some more time
• Gaussian 3 theories
  • G3 - improving on G2
  • G3(MP2) - again saving some time
  • G3B3 - using B3LYP geomtries
  • G3(MP2)B3 - the most economical method
  • G3X - aiming for even higher accuracy
• Gaussian 3 theories for open shell systems
  • G3-RAD - G3(B3) for open shell systems
  • G3X(MP2)-RAD - G3X(MP2) for open shell systems
  • G3(MP2)-RAD - G3(MP2)B3 for open shell systems
• Gaussian 4 theories
  • G4 - improving on G3X and optimizing for larger test sets
  • G4(MP2) - an economical variant of G4
• a note on relative performance
• further modifications of Gaussian theories
• a note on accuracy and precision