The rapid development of online news media requires websites to have high performance in order to deliver information quickly, stably, and optimally on search engines. The objective of this study is to analyze and compare the performance of the Next.js, Nuxt.js, and Remix.js frameworks in implementing Server Side Rendering (SSR) on news websites. The method used is a comparative experiment by building three identical news website prototypes using the three frameworks, then testing them with Google Lighthouse to measure rendering peformance and Apache JMeter to assess resilience to user load. The load testing was conducted in three scenarios using 250, 500, and 1000 virtual users/threads with a 5 second ramp-up period, a 300 second test duration, and an infinite loop count so that each virtual user continuously repeated requests until the test ended. Google Lighthouse test results show that Remix.js excels in the First Contentful Paint metric at 2.14 seconds, Largest Contentful Paint at 2.9 seconds, and Speed Index at 2.14 seconds, allowing news headlines and images to be displayed faster and improving user perception of speed and convenience. In JMeter testing, Nuxt.js showed the best performance with the lowest response time, highest throughput, and more stable error rate compared to other frameworks, demonstrating better resilience in the face of large traffic spikes. The conclusion of this study is that Remix.js is more optimal for improving user experience through content access speed, while Nuxt.js is superior in terms of scalability and reliability when traffic spikes occur. The implication of these findings is that the selection of an SSR framework for news websites should be tailored to the main priority, whether it is emphasizing access speed or system resilience under high load conditions.