Modeling geospatial tabular data with deep learning has become a promising alternative to traditional statistical and machine learning approaches. However, existing deep learning models often face challenges related to scalability and flexibility as datasets grow. To this end, this paper introduces Aggregator, an efficient and lightweight algorithm based on transformer architecture designed specifically for geospatial tabular data modeling. Aggregators explicitly account for spatial autocorrelation and spatial heterogeneity through Gaussian-biased local attention and global positional awareness. Additionally, we introduce a new attention mechanism that uses the Cartesian product to manage the model size while maintaining strong expressive power. We benchmark Aggregator against spatial statistical models, XGBoost, and several state-of-the-art geospatial deep learning methods using both synthetic and empirical geospatial datasets. The results demonstrate that Aggregators achieve the best or second-best performance compared to their competitors on nearly all datasets. Aggregator's efficiency is underscored by its reduced model size, making it both scalable and lightweight. Moreover, ablation experiments offer insights into the effectiveness of the Gaussian bias and Cartesian attention mechanism, providing recommendations for further optimizing the Aggregator's performance.