Micro gas turbine (MGT) technology is evolving toward a large variety of novel applications, such as weak gas electrification, inverted Brayton cycles, and fuel cell hybrid cycles; however, many of these systems show very different dynamic behaviors compared to conventional MGTs. In addition, some applications impose more stringent requirements on transient maneuvers, e.g., to limit temperature and pressure gradients in a fuel cell hybrid cycle. Besides providing operational safety, optimizing system dynamics to meet the variable power demand of modern energy markets is also of increasing significance. Numerical cycle simulation programs are crucial tools to analyze these dynamics without endangering the machines, and to meet the challenges of automatic control design. For these tasks, complete cycle simulations of transient maneuvers lasting several minutes need to be calculated. Moreover, sensitivity analysis and optimization of dynamic properties like automatic control systems require many simulation runs. To perform these calculations in an acceptable timeframe, simplified component models based on lumped volume or one-dimensional discretization schemes are necessary. The accuracy of these models can be further improved by parameter identification, as most novel applications are modifications of well-known MGT systems and rely on proven, characterized components. This paper introduces a modular in-house simulation tool written in fortran to simulate the dynamic behavior of conventional and novel gas turbine cycles. Thermodynamics, gas composition, heat transfer to the casing and surroundings, shaft rotation and control system dynamics as well as mass and heat storage are simulated together to account for their interactions. While the presented models preserve a high level of detail, they also enable calculation speeds up to five times faster than real-time. The simulation tool is explained in detail, including a description of all component models, coupling of the elements and the ODE solver. Finally, validation results of the simulator based on measurement data from the DLR Turbec T100 recuperated MGT test rig are presented, including cold start-up and shutdown maneuvers.