The paper delves into object detection with focus on You Only Look Once (YOLO) algorithm and its versions, i.e., YOLOv3, YOLOv5, YOLOv7, YOLOv8. Also studies the performance of the versions of YOLO by comparing them in terms of accuracy, speed, confidence, and their real-world applications. YOLOv3 displays impressive precision (0.79), YOLOv5 has performed consistently throughout in all cases. However, Yolov7 struggles to detect objects in scenarios of overlapping objects. YOLOv8 had difficulty in detecting objects in cluttered scenes. This comprehensive study guides researches, professionals, and curious students in selecting the most appropriate YOLO variant ideal for their task ranging from surveillance to autonomous driving.