The identification of Arabic characters and numbers on Iraqi license plates is essential for automated vehicle identification systems used in security, traffic control, and toll collection. Manual recognition of these characters can be time-consuming and error-prone, necessitating the development of an automated detection system. This study explores the effectiveness of You Only Look Once version 9 (YOLOv9) in recognizing Arabic characters on license plates, ensuring both accuracy and efficiency in vehicle identification. A specialized dataset featuring images of license plates with Arabic characters was used to evaluate YOLOv9’s performance. Metrics such as precision, recall rate, F1 score, and Mean Average Precision (mAP) were employed to assess the model’s success. YOLOv9 performed impressively across all metrics, including accuracy, recall, precision-confidence, and recall-confidence. The task, executed over ten epochs, took 28.5 min. In comparison to previous cutting-edge systems, YOLOv9 demonstrated superior performance in detecting Arabic letters on license plates, making it a highly effective solution for vehicle identification systems. Its balance of speed and accuracy makes it well-suited for practical applications, providing a reliable and efficient approach for automated vehicle recognition.