Animal models are widely used to develop and/or evaluate therapeutic approaches for humans. Additionally, kinematic gait analysis is conducted to determine the effectiveness of proposed therapies, such as those for the recovery of individuals with spinal cord injuries, based on angular kinematic parameters. These parameters can be used as feedback for real-time systems. With this in mind, we have developed Python software that extracts angular kinematic parameters from the hindlimbs in rats during real-time locomotor training. This extraction is based on the coordinates of anatomical landmarks identified by a DeepLabCut-trained neural network (DLC). The results obtained for the angular kinematic parameters were consistent with those of the literature, and the system latency approached the ideal, indicating that the software can assist researchers in real-time rodent kinematic analysis studies.