Most continuous-control-set (CCS) type model-free predictive controls (MFPCs) require to be implemented through intelligence functions. But due to different gain requirements in changing operating processes, the optimal gain selection in these functions is a major challenge. To address this issue, a Dahllin-based fast design scheme for intelligence functions is proposed in this paper, and applied into the CCS-type MFPC on permanent magnet synchornous motor (PMSM) drives. Using the Dahllin principle, both the MFPC and the plant are represented as an inertia term, allowing for the selection of optimal gains based on ideal responses. Additionally, boundaries are established through pole/zero mapping analysis to maintain system stability. The CCS-type model-free predictive current control using autoregressive with exogenous input (ARX) is selected as an example to test the proposed method, and its experimental results demonstrate the improvements in essential robustness and control quality of current.