We simulated the effects of the 21 August 2017 total solar eclipse on the ionosphere‐thermosphere system with the Global Ionosphere Thermosphere Model (GITM). The simulations demonstrate that the horizontal neutral wind modifies the eclipse‐induced reduction in total electron content (TEC), spreading it equatorward and westward of the eclipse path. The neutral wind also affects the neutral temperature and mass density responses through advection and the vertical wind modifies them further through adiabatic heating/cooling and compositional changes. The neutral temperature response lags behind totality by about 35 min, indicating an imbalance between heating and cooling processes during the eclipse, while the ion and electron temperature responses have almost no lag, indicating they are in quasi steady state. Simulated ion temperature and vertical drift responses are weaker than observed by the Millstone Hill Incoherent Scatter Radar, while simulated reductions in electron density and temperature are stronger. The model misses the observed posteclipse enhancement in electron density, which could be due to the lack of a plasmasphere in GITM. The simulated TEC response appears too weak compared to Global Positioning System TEC measurements, but this might be because the model does not include electron content above 550‐km altitude. The simulated response in the neutral wind after the eclipse is too weak compared to Fabry Perot interferometer observations in Cariri, Brazil, which suggests that GITM recovers too quickly after the eclipse. This could be related to GITM heating processes being too strong and electron densities being too high at low latitudes.