Forecasting solar radiation is essential for optimizing renewable energy systems by predicting the availability of solar energy. This study compares the performance of various machine-learning models for predicting solar radiation over a 240-h period in northern Morocco. The models assessed include Neural Networks (NN), Support Vector Regression (SVR) with Linear, Polynomial, and Radial Basis Function (RBF) kernels, as well as Basic Recurrent Neural Networks (RNN) and Long Short-Term Memory networks (LSTM). Performance metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and coefficient of determination (R2) were used to evaluate the accuracy of each model. The results show that LSTM outperformed the other models, achieving an R2 of 0.96, followed closely by SVR with an RBF kernel (R2 = 0.95) and SVR with a Polynomial kernel (R2 = 0.94). Neural Networks and Basic RNN achieved R2 values of 0.92 and 0.91, respectively, while SVR with a Linear kernel showed the lowest performance with an R2 of 0.85. These results highlight the effectiveness of LSTM and SVR models in accurately predicting solar radiation, providing valuable insights for renewable energy planning and grid management.