Relevance. The accumulation of information (statistical data and knowledge) about the COVID-19 pandemic leads to the refinement of mathematical models, to the expansion of the area of their use. The aim of this study is to build a set of models (in line with current knowledge and data) to identify the functions that drive the dynamics of a pandemic and analyze the possibilities for making predictions. Materials and methods. The work used data from open statistical and information resources relating to all aspects of COVID-19. The basis of the study is the balanced identification method and the information technology of the same name, created at the Center for Distributed Computing of the Institute for Information Transmission Problems of the Russian Academy of Sciences. The technology is used to build (select) models that correspond to the quantity and quality of data, perform calculations (forecasts) and present results (all the graphs below were prepared on its basis). Result. The constructed models satisfactorily describe the dynamics of the incidence of COVID-19 in Moscow. They can be used for a forecast with a horizon of several months, provided that new, previously absent elements do not appear in the modeled object. The main internal mechanism that determines the dynamics of the model is herd immunity and an increase in the infectivity of the virus (due to the spread of Delta and Omicron strains). Conclusion. The results of the successful use of balanced identification technology for monitoring the COVID-19 pandemic are presented: models corresponding to data available at various points in time (from March 2020 to December 2021); the acquired new knowledge - functional dependencies that determine the dynamics of the system; calculations of various epidemic indicators (morbidity, immunity, reproduction indices, etc.); various forecasts for Moscow (from 12/01/2020, 04/15/2021, 08/01/2021 and 08/01/2021).