The detachment and calculation of functionalities from a vehicle into a cloud creates new chances. By linking different data sources with the in-vehicle data in the cloud, an optimization of these functionalities in terms of en-ergy efficiency can be applied. For example, the Heating, Ventilation and Air Conditioning (HVAC) consumes up to 30% of total energy in a vehicle. Electric vehicles in particular lead to these high values because they are not able to re-cover the waste heat from combustion engines for interior heating. Therefore, the optimization of energy efficient strategies with respect to the vehicle energy management system becomes more relevant. Forecasts of the interior vehicle temperature are directly related to the HVAC energy consumption. This work focuses on the implementation and accuracy evaluation of Recurrent Neural Networks (RNN) for interior vehicle temperature forecasting.