The SWISS-MODEL workspace is a web-based platform for protein structure homology modelling. It provides an integrated environment for users to build homology models of proteins without needing to download or install large software packages. The workspace includes tools for template selection, model building, and structure quality evaluation. It allows users to access and update sequence and structure databases, and provides a personal working environment for multiple projects. The system supports three modelling modes: automated, alignment, and project, each requiring varying levels of user intervention. The automated mode is suitable for cases with high target-template similarity, while the project mode allows for manual adjustment of model building steps. The workspace also includes tools for domain assignment, secondary structure prediction, and disorder prediction. The SWISS-MODEL template library is derived from the Protein Data Bank and includes information for rapid retrieval of structural data. Template identification is performed using methods such as Blast, iterative profile Blast, and HMM-based searches. The system has been continuously validated through benchmarking and evaluation projects, demonstrating its accuracy and reliability. The SWISS-MODEL workspace has been used to model two human proteins, Cyclin A1 and Transmembrane Protease 3, showing its effectiveness in generating reliable 3D models. The system provides graphical outputs and quality assessments to help users evaluate model accuracy. The workspace is accessible via a web browser and is supported by a back-end system for data storage and processing. It is a valuable resource for comparative protein modelling, complementing other tools and databases in the field.The SWISS-MODEL workspace is a web-based platform for protein structure homology modelling. It provides an integrated environment for users to build homology models of proteins without needing to download or install large software packages. The workspace includes tools for template selection, model building, and structure quality evaluation. It allows users to access and update sequence and structure databases, and provides a personal working environment for multiple projects. The system supports three modelling modes: automated, alignment, and project, each requiring varying levels of user intervention. The automated mode is suitable for cases with high target-template similarity, while the project mode allows for manual adjustment of model building steps. The workspace also includes tools for domain assignment, secondary structure prediction, and disorder prediction. The SWISS-MODEL template library is derived from the Protein Data Bank and includes information for rapid retrieval of structural data. Template identification is performed using methods such as Blast, iterative profile Blast, and HMM-based searches. The system has been continuously validated through benchmarking and evaluation projects, demonstrating its accuracy and reliability. The SWISS-MODEL workspace has been used to model two human proteins, Cyclin A1 and Transmembrane Protease 3, showing its effectiveness in generating reliable 3D models. The system provides graphical outputs and quality assessments to help users evaluate model accuracy. The workspace is accessible via a web browser and is supported by a back-end system for data storage and processing. It is a valuable resource for comparative protein modelling, complementing other tools and databases in the field.