<p>This study addresses the clustering task of large-scale panel data with assistance from multi-network information under the grouped factor model. In many real-world clustering tasks, multiple networks can be observed for the same set of cross-sectional units based on different types of interactions. Different networks are different portraits of latent group memberships, which inspired us to utilize multi-network information to improve the clustering accuracy and stability. Therefore, we propose a multi-network-assisted clustering method that encourages coherence between the clustering results and the weighted multi-network in a penalized manner. We also developed a flexible weight learning strategy to identify the clustering capacity differences of multiple networks. A computationally efficient algorithm with random initialization was developed to implement penalized estimation. Thorough simulation studies demonstrate that the proposed method is more promising than existing competitors, even with misleading network information. Finally, application to the daily returns of stocks traded on the Shanghai and Shenzhen stock exchanges demonstrates the effectiveness and efficiency of the new method. Supplementary materials for this article are available online.</p>