The rapid growth of modern information systems, cloud computing, and data-intensive applications has increased the demand for database architectures capable of delivering high performance, scalability, reliability, and efficient resource management. This study presents a Systematic Literature Review (SLR) on the performance trade-offs among centralized, distributed, and cloud database systems. The research adopted a PRISMA-inspired Google Scholar Screening Workflow to identify, screen, and synthesize relevant studies published between 2018 and 2026. Literature selection was conducted using predefined keywords related to database performance, benchmarking, scalability, and architectural evaluation. A total of 29 selected studies were analyzed using qualitative synthesis methods focusing on evaluation metrics, architectural characteristics, benchmarking approaches, and performance outcomes. The findings indicate that centralized databases remain effective for environments requiring strong consistency, simplified administration, and controlled workloads, although they face limitations in scalability and fault tolerance. Distributed databases demonstrate superior performance in horizontal scalability, redundancy, and distributed workload processing but involve greater synchronization and transaction management complexity. Meanwhile, cloud databases provide deployment flexibility, elastic resource allocation, and service-based management models, while introducing trade-offs concerning operational cost, security, consistency, and deployment complexity. The review also identifies latency, throughput, response time, scalability, availability, and resource utilization as the most commonly used database performance metrics, with benchmarking frameworks such as Yahoo! Cloud Serving Benchmark (YCSB) playing a significant role in performance evaluation. The study concludes that no single database architecture is universally optimal, and database selection should be aligned with workload characteristics, business requirements, and deployment environments.