Identifying core single nucleotide polymorphisms (SNPs) associated with rice abiotic stress traits computationally helps researchers to deepen their understanding of genetic diversity, trait inheritance, yield improvements, and marker-assisted selection in rice breeding. The advancement of recent Artificial Intelligence (AI) can facilitate the identification and prediction of essential SNPs. In this study, a RiceAML model is proposed via utilizing Auto Machine Learning (AML) approach like (TPOT) to classify SNPs of abiotic stress traits in rice. Eight datasets (six individual datasets and two combined datasets) with two DNA sequence lengths 40 and 100 nucleotides, are used. Three encoding approaches are applied to convert DNA sequences into numeric values such as One-Hot, DNA2Vec, and concatenation of both feature vectors. The results showed that the prediction accuracy of the model trained on merged or combined dataset, especially Combined_100, outperformed the prediction accuracy of models trained on individual datasets, particularly when using the One-Hot Encoding, achieving an accuracy of 93.7%. The proposed RiceAML model with related datasets can be accessed through https://github.com/HasnaaImad/RiceAML .