<p>Cloud manufacturing service evaluation is crucial for selecting appropriate cloud services and enabling manufacturers to utilize extensive cloud resources efficiently. However, existing research relies on fixed evaluation indicators and subjective methods, limiting their ability to meet personalised evaluation requirements of service demanders. Therefore, a flexible evaluation method based on personalised evaluation requirements was proposed for cloud manufacturing services. Firstly, evaluation requirements from service demander groups for each cloud manufacturing resource type were collected, and their importance was calculated from usage preferences and frequency. A classification rule based on the importance of evaluation requirements was established to identify common and individual evaluation requirements per resource type. Secondly, the mapping between these requirements and existing evaluation indicators was analyzed to determine common and individual indicators per resource type. A flexible matching rule aligned tasks with appropriate indicators using the mapping between resources and tasks. Thirdly, a fixed-variable combination weighting method—integrating the CRITIC method, game theory, Spearman, and Kendall correlations—was developed to compute indicator weights by combining fixed and variable weights and assessing their consistency. Indicator values were quantified using historical manufacturing resource data, and service quality scores of candidate resources were calculated via weighted fusion of indicator weights. Eventually, the method’s effectiveness was demonstrated through a custom gear example and a case study of an automated guide vehicle manufacturing task, highlighting its utility in selecting cloud manufacturing resources that meet service demander requirements.</p>