Abstract: Deep learning algorithm are useful for investigation of ionospheric weather using past ionospheric data under various space weather conditions. The Total electron content (TEC) is an important parameter of the ionosphere and prediction of TEC are very challenging task, mainly in anomaly crest station. This research develops and analyzes a technique based on long short‐term memory (LSTM) neural network (NN) for the short-term ionospheric TEC prediction. In this work, multi-input LSTM forecasting technique is utilized and tested for evaluating its capability in prediction of ionspheric TEC over Lhasa, China station (Longitude: 91.10397200 , Latitude : 29.65734166) using vertical Total Electron Content (TECV) with Solar and Geomagnetic time series data. The results were then compared with the observed TEC collected by the IGS network. The result shows that the model could recognize the variation trend of typical TEC profile and have a good performance of the short term ionospheric TEC prediction.