Two apparatuses with lifting spherical source masses are built and used to verify the precision of gravimeters. The 333-kg source mass produces a maximum acceleration of 200 nm/s 2 with an uncertainty of 0.31 nm/s 2 , which corresponds to a relative uncertainty of 0.16%. After evaluating the temperature effect, drift of the gravimeter, the atmospheric effect, and the tidal effect, a combined uncertainty of 1 nm/s 2 is obtained. One CG6 gravimeter is tested using two apparatuses, the measured accelerations agree with the theoretical values within the error range. Differential measurement with two CG6 gravimeters on one apparatus is performed, which provides a common-mode rejection of the effects due to ambient noise, such as the gravity tide, atmospheric effect, and drift. The difference in acceleration measured by the two gravimeters is determined to be 199 ± 6 nm/s 2 , which agrees well with the value 200 ± 1 nm/s 2 obtained by using apparatus II. Our apparatuses provide a verification of the precision of gravimeters with an uncertainty of 1 nm/s 2 , which is one of the lowest uncertainties reached so far. The determination of geometrical metrology and mass distribution and detailed error analysis are presented. The methods on error analysis as well as differential measurement used in our work are helpful for gravity measurement.