The development of technology when making an explosion of the number of news or news documents that are very much on the internet, it is necessary to have clustering done in dividing these documents so that they can be adjusted based on the category of the online news. Application of document clustering can increase the effectiveness of information retrieval by referring to a hypothesis that relevant documents will tend to be in the same cluster if a collection of documents has been clustered. This research aims to try to do clustering on online news about Covid-19 taken from three online news websites contained in XML files. K-Means clustering is used as a grouping of online news by using an open-source application, Carrot2 Workbench which turns out to be able to generate nine clusters of "Covid-19" queries entered in the Carrot2 application.