Current methods for grasp planning are often constrained by predefined rules or manually designed action sequences, which limits their applicability and flexibility.In order to expand the applicable space of robotic arm grasping and improve the intelligence level of autonomous grasping by robotic arms,explored an AIGC-based automatic generation method for robotic arm grabbing action sequences.This method is centered around the open-source Chinese large language model ChatGLM2, utilizing the generation capabilities of large language models to automatically generate sequences of robotic arm movements and create executable programs.To verify the proposed method, the UR5 robotic arm was selected as the experimental subject, and a large number of experiments were conducted in the Coppelisim simulation environment,verify whether the generated sequence program of grasping action can successfully complete the grasping task,the effectiveness of the proposed method was evaluated.Experimental results show that the automatic generation method based on AIGC can effectively generate the mechanical arm's grasping action sequence and executable program that meet the experimental goals.