Gene fusions drive oncogenesis and therapeutic resistance in hematologic malignancies, with the Philadelphia chromosome BCR-ABL1 fusion serving as a paradigmatic target in chronic myeloid leukemia. Current clinical fusion detection relies on bulk RNA sequencing that masks cellular heterogeneity and requires 5–7 days for results, limiting real-time therapeutic decisions. Short-read sequencing cannot span complete fusion breakpoints, while bulk approaches obscure rare fusion-positive cells contributing to treatment resistance. We developed a rapid single-cell long-read RNA sequencing pipeline integrating ArgenTag single-cell capture, long-read sequencing, and CTAT-LR fusion detection to characterize Philadelphia chromosome heterogeneity in K562 cells. Our workflow employs a one-shot mathematical demultiplexing algorithm that assigns cellular identity directly from barcode sequences without external references. Analysis of 7,856 single cells revealed 910 unique fusion events, with BCR-ABL1 detected in only 0.23% of cells (18/7,856), demonstrating significant fusion heterogeneity previously masked by bulk approaches. Fusion-positive cells distributed uniformly across cell cycle phases (11.9–17.5% per cluster), indicating state-independent expression. Molecular characterization precisely mapped the canonical t(9;22) breakpoint and confirmed BCR-ABL1 transcript architecture at single-cell resolution. Our complete workflow achieves fusion detection and clinical reporting in <24 h compared to 5–7 days for conventional methods, enabling real-time precision oncology applications. This approach transitions cancer genomics from population-averaged to individual cellular insights, revealing clinically relevant fusion heterogeneity that may guide personalized therapeutic strategies and resistance monitoring.