Machine Learning helps in identifying patterns in ciphertexts. Classification of the cipher is a crucial task for cryptanalysis and this serves as the initial step before proceeding the decryption. It is an important step in cryptanalysis, and thus needs ease and accuracy to be able to produce correct results. To address this, we investigate classical ciphers classification employing Machine Learning technique on ROT13, Caesar, Vigenère, Hill, Columnar Transposition, Substitution and Affine ciphers. For this, an efficient Support Vector Machine (SVM) model is applied on these ciphers to identify the encryption algorithm. Seven types of ciphertexts are generated corresponding to the plaintext of size 979 KB. Evaluation metrics are used to assess the model’s performance, ensuring its reliability and accuracy. The SVM model achieved an average accuracy of 99.56%, with its highest accuracy of 99.95% recorded for the Caesar cipher, followed by the Affine and ROT13 ciphers.