Grassland ecosystem is one of the most important and widely distributed ecosystem types in terrestrial ecosystems. Grassland biomass is an important indicator to characterize and evaluate grassland ecosystems. The purpose of this paper is to study a typical grassland biomass model based on vegetation index. Grassland A is used as the research area, the research object is grassland aboveground biomass, and the aboveground biomass data of field measured sampling points are used as modeling data. Based on the NDVI vegetation index, the correlation analysis between RVI and the measured value of above-ground biomass in grasslands, respectively, established a linear model, an exponential model and a quadratic polynomial model. Through SPSS statistical analysis, the accuracy of the model was tested, and the aboveground biomass of remote sensing grassland was inverted using the optimal model. The research results show that the aboveground grassland biomass simulated by the power function model with RVI-Y=102.45x1.35 is basically close to the measured aboveground biomass. The average error of the estimated error/measured value is 6.79%, and the fitting accuracy is 94.6%. Therefore, it is feasible to estimate grassland biomass using an exponential function model based on RVI-above-ground biomass.