Abstract: The ensemble of the more than 6,000 known extrasolar planets in our galaxy exhibits a diversity of objects unseen in our solar system. With space missions like the Hubble Space Telescope, CHEOPS, the James-Webb Space Telescope and soon also PLATO, exoplanet research has turned from discovering such planets outside the solar system to characterizing these objects that orbit stars other than our Sun.  Recent progress in characterizing exoplanets with the CHEOPS, TESS, and JWST space telescopes has demonstrated that extrasolar gas giants exhibit climate regimes ranging from ultra-hot to rather cold. Ultra-hot Jupiters (like WASP-18b and Kelt-7b) exhibit enormous day-night temperature different of more than 2000K. Strong eastward wind jets can therefore transport hot gases from the dayside to the nightside where mineral cloud particles form.  In order to physically interpret such complex atmosphere behavior, our research group has built virtual laboratories that enable us to predict temperature, velocity, cloud and chemistry  maps for planets orbiting different host stars. Such virtual laboratories that describe and explore the atmosphere structure and cloud formation allow us detailed studies of specific planets (e.g. the JWST target WASP-39b) as well as ensemble studies, for example,  in preparation of observational campaigns. Machine learning techniques are helping to fill gaps where deterministic modelling is too time-consuming.