The molding method is highly anticipated in engineering fields and has been preliminarily applied. However, traditional demolding methods introduce subtle deviations in demolding direction, leading to demolding defects and reduced forming accuracy. To address this, we developed an innovative Shape Memory Polymer (SMP) mold. Through finite element simulation, we extensively investigated the effects of diaminodiphenylmethane (DDM) addition, temperature, and load on the template forming and shape recovery processes. Finally, we experimentally verified the feasibility of the self-demolding of the template. The findings demonstrate that as the amount of DDM added increases, the glass transition temperature of SMP progressively elevates. This can be attributed to the enhanced crosslinking triggered by the addition of DDM. Consequently, during the stage of preserving shape in mold formation, SMP exhibits a heightened storage modulus and diminished shape recovery. Notably, when 1.5 g of DDM is added, the shape memory template exhibits the least shape recovery, with a shape retention rate of 78.2%. Conversely, employing 1.6 g of DDM generates the highest shape recovery but only achieves a shape retention rate of 59.5%. Augmenting the amount of DDM is advantageous in accomplishing template demolding. Higher temperatures expedite the initiation of the shape recovery process, facilitating template demolding during the shape recovery stage. Moreover, increasing the load can minimize template rebound and enhance the precision of mold formation. However, it also intensifies the complexity of self-demolding. Consequently, it is imperative to sensibly determine and implement suitable process parameters in applications.