Lung cancer remains the leading cause of cancer-related mortality worldwide, largely due to therapeutic resistance and tumor recurrence. In non-small cell lung cancer (NSCLC), cancer stem cells (CSCs) have emerged as key contributors to tumor persistence and progression. However, accelerated CSC-targeted drug discovery is limited by the lack of scalable and quantitative screening systems capable of capturing stemness-associated phenotypes. Conventional two-dimensional (2D) assays often fail to reflect the architectural and molecular complexity of CSC-enriched tumor populations. Here, we report the establishment of a luciferase-based three-dimensional (3D) NSCLC spheroid platform integrating CSC-like enrichment with a high-throughput quantitative readout. The luminescent signal strongly correlated with spheroid burden and viable cell content. Compared with 2D monolayers, 3D NSCLC spheroids exhibited elevated expression of pluripotency-associated factors, including OCT4, NANOG, and SOX2, and CSC-associated markers, namely CD44, ALDH1A1, and ABCG2, supporting the enrichment of CSC-like subpopulations. As a proof-of-concept, screening of renieramycin T right-half derivatives identified DH_18 as a lead compound in the established 3D NSCLC spheroid model, exhibiting greater potency than DH_19, accompanied by increased apoptosis and reduced CSC-associated proteins in follow-up experimental validation. Integrated network pharmacology incorporating enrichment and protein–protein interaction analyses further mapped DH_18–NSCLC targets to lung CSC-associated networks and pathways related to stemness and aggressive tumor phenotypes, reinforcing the biological relevance of the 3D screening context. Collectively, these findings present a quantitative and scalable 3D NSCLC spheroid platform suitable for CSC-targeted drug screening and nominate DH_18 as a candidate compound with potential activity against stemness-linked molecular networks.