In the field of network science, identifying influential spreaders in complex networks is a highly concerned research topic. Existing approaches frequently assign identical values to a substantial number of nodes, resulting in lower differentiation among these nodes. To tackle the issue of inadequate resolution in traditional centrality measures, we propose a novel method for identifying influential nodes \(-CE\) -based gravity model (CEGM)−which integrates the extended degree, an improved E-shell, and a gravity model to evaluate the impact of nodes in an all-around manner. It aims at offering a more comprehensive measure of the node’s influence than the known methods. We conduct comprehensive experiments involving CEGM and nine benchmark methods, utilizing six real-world datasets, and evaluate metrics including the SIR model, Kendall’s correlation coefficient and monotonicity index. The experimental results demonstrate that, our method CEGM can not only achieve a comparable or even better recognition performance but also obtain a much fairer and better rank result compared with current algorithms, highlighting the effectiveness and superiority of our proposed scheme.