The recent new normal occasioned by the outbreak of Covid-19 pandemic has brought online distance education system to limelight because it is one major protocol to containing the impact and spread of the deadly disease since it completely averts physical gathering. In this trend, monitoring and predicting learner’s academic performance (LAP) becomes highly imperative so as to pay adequate attention to pertinent areas where necessary adjustments should be made concerning the learner’s academic progression. Unlike the conventional mode of education, doing these is herculean task, therefore, this study developed a system that can efficiently predict LAP through Adaptive Neuro-Fuzzy Inference System (ANFIS) implemented on MATLAB 18.0 interface. Following the standard procedures, the structure of ANFIS model was optimized and utilized to predict the Cumulative Grade Point Average (CGPA) of learners. ANFIS-models were further used to investigate the interactions among the input variables and the corresponding effects on outputs, CGPA. Results showed accuracies of ANFIS-models based on the values of root-mean-square-errors (RMSE). Values obtained for RMSE ranged from 0.7999 to 28.6560 for all the ANFIS-models developed while the optimal model, ANFIS-Gaussmf, accurately predicted the LAP because it had the least RMSE value of 0.7999 at 100 epochs. Also, CGPA was found to depend majorly on the varying inputs in forms of scores in examination as direct proportional relationship was observed. T-test conducted to determine any significant difference between the collected and predicted data recorded p-value of 0.084, thus, signifying efficiency of the developed model. This study has established ANFIS-model to be a very reliable tool for prediction of LAP so as to ensure excellent results even on completion of study.