In order to scientifically and reasonably select the third-party service organizations of equipment price, the evaluation model was constructed by gray correlation analysis and neural network.The research data was obtained in the form of questionnaire survey. The gray correlation method was used to analyze the correlation between the evaluation index and the comprehensive score. It is found that the two primary evaluation indicators of service attitude and system construction are more important, with a correlation degree of more than 0.74. Using effective secondary indicators and comprehensive scores as input and output data, the evaluation model based on neural network was built. Finally, through the verification and analysis of the rationality of the method, it is found that the prediction accuracy probability of the neural network model is 75% and the probability of meeting the requirements is 25%, which meets the actual user needs and can provide suggestions for the competent departments to select third-party institutions.