With the development of smart mobile computer, more and more portable devices are coming into market that includes wearable devices. To evaluate the wearable devices, the wearable level is an important reference. In this paper, we proposed to evaluate wearable devices based on quality characteristics. However, the characteristics are usually unstable, historical sample size is small, index, show obvious non-normal distribution. These features make it hard to use the general statistical techniques for effective evaluation. A model based on support vector machine (SVM) is proposed to distinguish the wearable level. This paper introduces the current situation of wearable devices, descripts its classification and evaluation principle, and gives out the classification model and algorithm for wearable level. This work can be a beneficial reference to wearable standard.