Force-based and displacement-based approaches in seismic codes have been widely used in structural design practices for decades. These methods use either the peak forces or the maximum deformations to quantify the destructive potential of a structure subjected to ground motion and does not account for either the duration of ground motion or the loading history on the structure. Hence, several studies proposed many energy-based parameters, of which, the equivalent velocity (VEA) considered as a stable measure and an effective tool in energy-based seismic design. As a preliminary step towards energy-based design in India, we propose design energy spectra for the Peninsular India (PI) in this study as it can provide an improved means of framework in the earthquake resistant design. In order to quantify the probability of exceedance in probabilistic seismic hazard analysis (PSHA), a ground motion model (GMM) for VEA is derived for the PI region using an Artificial Neural Network (ANN) technique. Since the ground motion data is limited in the PI shield, records from a similar tectonic region, such as the Central and Eastern America (CENA) database, are used for deriving the GMM. In the GMM, magnitude (Mw), rupture distance (Rrup), average shear wave velocity of the soil in the top 30 m (Vs30), and focal mechanism, are used as predictor variables; and the target variable constitutes the logarithm of VEA at 24 spectral periods. Consequently, uniform hazard energy response spectra (UHERS) for a 2475-year return period are obtained. In addition, acceleration spectra corresponding to the resulting UHERS are compared with that of the Indian Standard (IS) code recommended design spectra to observe for any underestimations. Finally, a tripartite Design Energy Spectra (DES) is proposed for the entire PI shield, for ease of access.