<p>For the mobile robot to be autonomous, the basic requirement is that the mobile robot should be able to localize itself in the environment and know the limitations of the environment. These will help the mobile robot plan the environment’s paths and avoid obstacles. For path planning, sample-based algorithms have been widely used. The most commonly used algorithm is RRT*, as it provides asymptotic optimality. As the sampling-based algorithms are in mobile robots for path planning, the problem arises when these algorithms provide paths which is not smooth and also close to the obstacle. To solve this problem, Beast RRT* is introduced. This paper proposes a new algorithm, Beast RRT*, which provides a better solution than RRT*. In this paper, the experiment is done with 1000 different types of environments in both two-dimensional and three-dimensional environments. The aim of this paper is to provide a new RRT* variant algorithm that can be used in mobile robots for real-time application, as the main function of Beast RRT* is to enhance path planning and provide planned paths in real-time. Beast RRT* functions in such a way that the samples generated are biased towards the goal, and also, magnetic force is added to the goal, which pulls the samples towards the goal. Due to smart sampling, the convergence rate of the RRT* is very low, and at the same time, the sample generated is also minimal. The effectiveness of Beast RRT* is demonstrative by comparing the algorithms to the existing algorithms through simulation.</p>