In the growing energy demand, the renewable energy power generation plays an important role. The most important factor in solar PV system is the estimation of maximum power point calculation. This paper presents Neural and Neuro-fuzzy system-based temperature and solar radiation forecast. In this system the feature values are predicted without the knowledge of the characteristics of the input time-series. Using standard meteorological parameters, the input data used to train the proposed systems with different architectures. After having simulated, many different structures of neural networks are trained using measurements as training data. The best structures are selected in order to evaluate their performance. ANFIS neuro fuzzy system is considered here because it combines fuzzy logic and neural network techniques to get more gain and efficiency which gives best accuracy in performance. Several Error Metrics are considered here to evaluate and compare both the systems according to the resultant predictions.