The aim of this study is to design a model for collaborative learning in virtual environments, specifically focusing on the Islamic Azad Universities in Tehran. The research is applied in terms of its goal, employs qualitative data, and is exploratory in nature, based on a grounded theory approach. The study population consists of experts, and 15 experts were selected for interviews through theoretical sampling of the snowball type, based on theoretical data saturation. A semi-structured interview was used as the research tool, organizing the dimensions, components, and indicators of the collaborative learning model in virtual environments. Initially, several models, findings, studies, and national and global theories were reviewed to create the interview form. Subsequently, open coding was used to identify indicators, and axial coding categorized them into dimensions, components, and indicators, which were then sent to the 15 experts. Through selective coding, interviews, the Delphi method, and brainstorming, the dimensions and components were identified. As a result, three dimensions, nine components, and 62 indicators for the virtual collaborative learning model were finalized. Following the experts' final validation and prioritization, the model's dimensions, components, and indicators were mapped, and the model was re-validated by experts.