Part of speech tagging (POS tagging) is a part of Natural Process Language (NLP). POS tagging is the process of automatic labeling of a word in a sentence according to the word class. There are various tagger methods in POS tagging, each tagger method has its own characteristics in its application. The research method used is Conditional Random Fields and Hidden Markov Model. The training of the two method models uses the Indonesian language corpus and Indonesian news texts as test data to determine which method is the most efficient based on the results of the accuracy and training time of each model. The method that has the best value is the CRF method with an accuracy value of 97.68 on the evaluation of the corpus test data and 90.02% for the sample Indonesian news dataset with a training time of 146.90 seconds, then there is the HMM method which has the highest accuracy value with a value of 94.25 % and shorter training time relatively shorter at 32.45 seconds and for the sample sentences containing 116 tokens, CRF method produces 90.05% accuracy which is higher than the HMM method which produces 79.31% accuracy.