An industrial robot kinematic model with joint geometric error parameters and a calibration algorithm is proposed. Firstly, based on the DH model, six geometric error parameters are introduced for each joint to establish a more comprehensive error calibration model. The solutions of forward and inverse kinematic for the model are realized. Then, a differential kinematic Jacobian matrix containing 45 parameters, including joint errors, base coordinate errors, and tool coordinate errors is established. An iterative algorithm based on a small sample test set is used to solve the error parameters. Finally, experimental verification is carried out using a laser tracker on the SIASUN SR10C robot. The calibrated error parameters are then compensated into the model. Results show that, after calibration compensation, the maximum position error of the robot decreases by approximately 80%, the average position error decreases by approximately 80%, and the error variance decreases by approximately 97%, demonstrating that this method can significantly improve the absolute position accuracy and determinacy of industrial robots.