<![CDATA[The increasing demand for web-based applications capable of handling a large number of concurrent users requires the selection of backend technologies that provide optimal performance and resource efficiency. However, previous studies generally employed different configurations and testing environments, making performance comparisons among backend technologies less objective and difficult to generalize. This study aims to analyze and compare the performance of backend REST APIs developed using Node.js, .NET, and Laravel Octane within a standardized testing environment. The research adopted a quantitative experimental approach using load testing with Grafana K6 by simulating workloads of up to 150 virtual users through HTTP GET requests. The evaluated performance metrics included throughput or requests per second (RPS), response time (latency), error rate, as well as CPU and memory utilization. The results showed that Node.js achieved the best performance with a throughput of 493.93 requests per second and an average latency of 77.43 ms, followed by .NET with a throughput of 324.94 requests per second and a latency of 163.06 ms, while Laravel Octane achieved a throughput of 192.29 requests per second with a latency of 358.62 ms. All backend technologies maintained a 0% error rate throughout the testing process. In terms of resource efficiency, Node.js also demonstrated the lowest CPU and memory utilization compared to .NET and Laravel Octane. These findings indicate that differences in execution models significantly influence throughput, response time, and resource utilization efficiency, providing valuable insights for selecting appropriate backend technologies to develop high-performance and highly scalable web applications.]]>
