Many observational data will be available through Virtual Observatories, but at some wavelengths and in some directions, no data yet exist. Concerning star counts, photometry and classifications, galactic models can be usefully included in VO for statistical inputs. The Besancon model of population synthesis well calibrated at visible and NIR wavelengths can be useful to complement observational data: (i) to ex- trapolate the stellar statistics to fainter magnitudes, at new wavelengths or interpolate at directions where no data are available yet; (ii) to compute bayesian classification probabilities. The revised version of the model as well as the classification scheme are briefly described here.