The Sake Robotics Gripper is a cheap, robust and versatile gripper that has not been simulated yet. It is an underactuated gripper and therefore needs a workaround to achieve realistic behavior in simulation. The model has to be able to interpret the exact same ROS messages the gripper receives. This paper proposes a reproduction of the Sake Robotics Gripper in V-REP. We analyze the tools provided by V-REP to develop an algorithm to achieve underactuation in simulation. This contributes to the kitchen cleaning task of the GOAL-Robots project with a platform to do machine learning on. All models are available open source. Our model can be used as a foundation for research in complex grasping and manipulation tasks with the Sake Robotics Gripper