Youtube is one of the media social platforms that are the most used in 2021. Social Media is not only for uploading entertainment videos but source information. With covid - 19 pandemic, this platform gives us much information about the pandemic, new variant information, and some vaccines that are distributed to the public. Responses about vaccines are very diverse from the public, from positif to negatif responses which are commonly referred to as sentiments. However, to know the contents of the response, an analysis is needed to understand the sentiment better. This study aims to determine the performance of the Support Vector Machine algorithm in classifying based on comments from the Indonesian people. The data source is taken from the Youtube site by utilizing the existing comment feature. The model test was carried out using the python library, namely SVC. Three scenarios were carried out on the test data, namely the distribution of training data and test data of 90:10, 80:20, and 70:30 which were carried out randomly on the data. The test was carried out 10 times on each. The evaluation was carried out using Cross-Validation and each obtained an accuracy of 86 %, 85 %, and 84 %.