The upcoming 2024 Indonesian General Election (Pemilu 2024) will be interesting news for online media users. With so much news about the election, online media has become one of the most effective media used to guide public opinion. Apart from that, public opinion is that the coverage in online media for each candidate is not balanced or not because a media is considered to have an affiliation with a particular candidate. To prove this opinion, sentiment analysis will be carried out on several online media in order to prove whether people's opinions are correct or not. Although previous research has used various platforms and achieved various levels of accuracy using the Convolutional Neural Network (CNN) and Support Vector Machine (SVM) methods with various features, this analysis will be developed using the Convolutional Neural Network (CNN) method to obtain higher accuracy and will be compared with the Support Vector Machine (SVM) method from the media platforms Detik.com, CNN Indonesia and CNBC Indonesia. The final results prove that the use of the Convolutional Neural Network (CNN) method shows an average combined performance of 65% (Cancidate 1 = 61%, Candidate 2 = 69%, Candidate 3 = 65%) smaller than the performance of the Support Vector Machine (SVM) method. with a combined average of 74% (Candidate 1 = 73%, Candidate Candidate 2 = 77%, Candidate Candidate 3 = 72%). This study provides insights into optimizing sentiment classification techniques for Online Media platforms, emphasizing the importance of leveraging semantic and contextual information in sentiment analysis tasks.