Comparative analysis of architectural approaches to REST API development for high-load systems

In the context of increasing system loads and restricted access to foreign cloud services, the selection of an efficient REST API architecture becomes particularly critical. This study presents a comparative analysis of monolithic and microservice-based API implementations developed using modern frameworks: FastAPI, Django REST Framework, Laravel, ASP.NET Core, NestJS, Symfony, and Spring Boot. All implementations followed a unified OpenAPI contract and were deployed in a self-hosted environment. Load testing was conducted using k6 and JMeter, with metrics including average and p95 latency, throughput, error rates, and resource utilization. The results indicate that FastAPI and ASP.NET Core provide optimal performance and resilience. Practical conclusions are drawn regarding the applicability of architectural models under high load and limited infrastructure.