This paper presents MultilabelDTree, a web-based application that uses automated machine learning for multilabel problems. The application allows users to upload their own datasets and build multi-label classification models by using their preferred problem transformation method along with optimized decision tree parameters. Model performance evaluation is conducted through k-fold cross-validation, providing detailed metrics for each label and an overall assessment. Users can save the pre-trained models for future use, visualize decision trees, and make predictions on instances of unclassified datasets. Additionally, the application offers an “Auto Mode” that automatically selects the best problem transformation method and parameter values based on the highest accuracy achieved. The usability of MultilabelDTree was assessed using the System Usability Scale. The results show its potential as a valuable tool in multi-label classification scenarios.