Information is one of the important things in this era, one of the information that always exists every day is news. The amount of news that appears every day becomes a new problem when news websites do not provide API (Application Programming Interface) services to get the news. This is an obstacle for researchers who will analyze news topics. The copy and paste method is less effective in getting news every day on news websites because it takes a long time. In this research, web scraping is done with the HTML (Hypertext Markup Language) DOM (Document Object Model) method to retrieve data from news sites. The results of web scraping are in the form of datasets which are then entered into the database and made into an API. The API that has been created is tested using black box testing and testing the suitability of the data, between the data obtained when scraping and the data on the news website at the time of testing. The results of testing using black box testing show that the filters on the API created run according to their functions and get a high percentage of data conformity. The Tribunnews.com news website has a conformity rate of 99.2%, Detik.com of 97.9% and Li-putan6.com of 98.6%.