Abstract: Based on a systematic review of the evolution and limitations of bibliographic systems, this paper proposes a next-generation catalog architecture for the era of generative artificial intelligence—Generative Public Access Catalog(GPAC). Centered on the core logic of “Integration-Understanding-Generation-Service,” GPAC integrates large language models, semantic retrieval, knowledge graphs, intelligent metadata, and verifiable generation mechanisms to create a new paradigm of intelligent catalog system based on natural language interaction, semantic intent parsing, and knowledge generation. Importantly, OPAC has not been replaced but rather embedded within GPAC as a foundational capability, serving as the underlying infrastructure supporting semantic computation and intelligent services. This paper focuses on the adaptation pathways and feasibility of GPAC in key library service scenarios, including the reconstruction of semantic search and recommendation services, the optimization of knowledge organization driven by intelligent metadata, and resource acquisition mechanisms oriented toward knowledge structures. The study demonstrates that GPAC is not merely a continuation of technological evolution of catalog systems but also a vital foundation for reimagining the philosophy and functions of library knowledge services. It offers both a theoretical framework and practical direction for building the next generation of intelligent, trustworthy, and collaborative public knowledge infrastructure.