The rapid and accurate identification of fungal pathogens remains a significant challenge in clinical microbiology. Surface-enhanced Raman spectroscopy (SERS) shows great potential for fungal diagnostics; however, achieving fast, convenient, and reproducible spectral acquisition remains difficult. Additionally, the high biochemical similarity among fungal species often leads to overlapping spectral features, limiting diagnostic accuracy. To overcome these limitations, we propose a temperature-regulated SERS spectral fusion strategy for the precise differentiation of clinically relevant fungal pathogens. Systematic profiling revealed that thermal modulation of nutrient-deprived fungi reprograms purine-associated metabolic outputs, producing distinct, temperature-specific spectral features. By integrating spectra acquired at 37, 45, 60, and 70 °C using a data-level fusion approach and applying a SoftMax-based classification model, we achieved 100% identification accuracy across both laboratory strains and clinical isolates. Notably, the thermal activation protocol significantly accelerated the SERS acquisition process, reducing the incubation time required to generate robust spectral signals from over 22 h at room temperature to just 1 h. Furthermore, the method eliminated the need for cell wall disruption, greatly simplifying sample preparation. Together, these findings underscore the potential of temperature-dependent SERS spectral fusion as a powerful tool for clinical fungal diagnostics and establish a broadly applicable, stimulus-responsive metabolomic fusion framework for scalable and precise pathogen identification.