Abstract: The timeliness and accuracy of earthquake emergency response are directly critical to the success rate of life rescue and the effectiveness of disaster loss control. Focusing on the core decision-making needs during the golden rescue window of 0–10 hours after an earthquake, this paper takes the M_S 6.8 earthquake that struck Dingri, Xizang, China, on 7 January 2025 as a typical case. It concentrates on the central assessment target—assessment of building damage severity and delineation of priority rescue areas—and systematically compares the applicability and performance differences of multi‑source heterogeneous data in rapid earthquake damage assessment. The study analyses the timeliness‑accuracy trade‑off characteristics of four types of techniques: seismic intensity attenuation models, instrumental intensity rapid reporting, mobile‑phone‑based population heatmap data, and interpretation of optical remote sensing imagery. On this basis, a "phased progressive" multi‑modal collaborative assessment framework is proposed. The results show that the comprehensive assessment outcomes derived from synergistic multi‑source data fusion outperform any single method in terms of spatial consistency, intensity identification accuracy, and hit rate of priority rescue areas, thereby validating the effectiveness of the collaborative framework. This study clarifies the applicability boundaries and prerequisites of various techniques in the post‑earthquake rapid assessment process, providing a technical reference for improving the scientific basis of decision‑making and the efficiency of rescue operations during the golden rescue period.