Infrared small target detection is a critical task in national defense, homeland security, aerospace, industrial monitoring, and environmental conservation, while the detection accuracy is generally insufficient due to the small size of the targets and the unclear features. Additionally, conventional detection algorithms often face challenges in terms of computational resources, making it difficult to achieve higher frame rates for infrared small target detection. To overcome these challenges, a new algorithm based on the latest YOLOv8 is introduced. In the feature extraction stage, the lightweight repeated C2f module from YOLOv8 is utilized, along with the incorporation of an attention mechanism. This approach reduces computational overhead while enhancing feature extraction capabilities. The detection results on the DIRST dataset demonstrate that the proposed method achieves an 8.4% increase in frame rate compared to the baseline network. Furthermore, it also exhibits improved performance in the confusion matrix analysis.