AbstractGlobal Ionosphere Thermosphere Model (GITM) results have been compared with measurements from Global‐scale observations of the limb and disk (GOLD) and Sounding of the Atmosphere using Broadband Emission Radiometry (SABER). For the first time, GOLD‐derived exospheric temperature and column‐integrated ratio measurements have been used to validate GITM model results. We examine two geomagnetic storm events for which we drive GITM with space weather conditions to understand how well the model reproduces the thermospheric responses to geomagnetic activity. In this paper, a recently developed auroral model, the Feature Tracking of Aurora (FTA) model, has been employed to calculate auroral electron precipitation in GITM (GITM w/FTA), and results are compared with the OVATION prime (OP) driven GITM model (GITM w/OP). GITM w/FTA simulated temperature, neutral density, and nitric oxide (NO) density are generally higher compared to the GITM w/OP model. During the geomagnetic storm, the GITM model and GOLD‐derived exospheric temperature agree between and N latitude in the equatorial region. GOLD measurements show strong peaks on either side of the equator during the geomagnetic storm period, which is also observed in our model results. The NO cooling peaks estimated by GITM models are 20 km lower than SABER observations during geomagnetic storms. GITM w/FTA better matches SABER observations than GITM w/OP except at low latitudes during storm time. Our model‐model/data comparisons show that improved auroral models are needed to better capture the thermospheric variations during geomagnetic storms.