Commercial banks are the main body of the finance industry in China. It is of great significance to study the impact of commercial banks’ spatial agglomeration on PM2.5 for China to develop a green economy. This article selects data from 30 provinces in China, covering 2000 to 2021. This study innovatively utilizes commercial banking institutions’ longitude and latitude geographic coordinate information to build a new indicator to characterize the spatial agglomeration degree of commercial banks. Then, we use the geographically and temporally weighted regression model to investigate the spatio-temporal heterogeneous effect of commercial bank agglomeration on PM2.5. The theoretical mechanism concludes that financial agglomeration exacerbates PM2.5 pollution through the scale effect and can also reduce PM2.5 pollution through technique effect and composition effect. Financial agglomeration and PM2.5 have obvious temporal and spatial differences as well as spatial autocorrelation characteristics. The geographically and temporally weighted regression model's results show that from a national perspective, financial agglomeration can inhibit PM2.5 pollution, but the inhibitory effect is gradually diminishing, indicating that it is imminent for China to further deepen its green financial reform. From the provincial level, the influence of financial agglomeration on PM2.5 has obvious temporal and spatial differences. The inhibitory effects of Beijing, Tianjin, and Hebei are becoming stronger, and these areas have the best situations. The promoting effects of the three northeastern provinces and Shanxi and other central and western provinces are becoming larger and larger, and these areas have the worst situations. Shanghai and other eastern provinces and Guangxi and other western provinces have respectively brought inhibitory effects and promoting effects, but the effects are all weakening, and the situations are in the middle. The scientific value of this study lies in the following: First, this study combines the environmental Kuznets curve theory for mechanism analysis, providing a scientific theoretical basis for subsequent related research. Second, the financial agglomeration index constructed in this study provides a scientific reference for academic circles to more accurately investigate the relationship between financial agglomeration and environmental pollution. Third, this study reveals the temporal and spatial differences in the impact of financial agglomeration on PM2.5 pollution by using the geographically and temporally weighted regression model for the first time, pointing out the focus and direction for decoupling economic growth and PM2.5 pollution under the influence of financial agglomeration in China provinces. With China's efforts to achieve green sustainable development, this study provides new ideas and valuable insights into the driving factors of green economic growth in China.