Accurate hydrological modeling is essential for water resource management in Central Asia, where meltwater from snow and glaciers support approximately 136 million people. Precipitation uncertainties significantly impact the accuracy of hydrological models, yet the scarcity of in-situ meteorological observations in the high mountain regions of Central Asia has led to widespread reliance on gridded precipitation products. These gridded products vary in quality and often require correction before use in hydrological modeling. Inverse hydrological modeling, where observed outputs (like river discharge) are used to infer unknown inputs (like precipitation), has been proven effective in estimating long-term average precipitation volumes, but accurately capturing interannual precipitation variability remains an unresolved challenge. Here, we develop an inverse modeling framework that integrates total discharge, baseflow, snow cover fraction, and glacier mass balance to simultaneously constrain both precipitation quantities and interannual variability across Central Asia’s river basins. A key innovation of this approach is the use of baseflow to constrain annual precipitation variability. By leveraging the strong relationship between winter snowpack accumulation and subsequent baseflow, this method enables a more accurate representation of interannual precipitation dynamics in the region. This work yields a benchmark precipitation dataset for hydrological modeling applications while also providing a systematic assessment of runoff regimes throughout this critical region.