Using YOLO v8, a cutting-edge deep learning model well-known for its speed and precision in object detection tasks, we present a unique method for real-time forest fire detection in this study. By utilizing a meticulously selected dataset that included a range of photos of forests with different levels of fire intensity, we were able to precisely train the YOLO v8 model using D-Fire dataset to identify and locate fire incidents. The methodology we employed included of extensive preprocessing of the data, such as augmentation and normalization, and then a thorough training and fine-tuning of the model. The evaluation findings show that our approach is effective, with noteworthy metrics for accuracy, precision, recall, and F1 score. By providing a workable method to protect ecosystems and human lives from the catastrophic effects of wildfires, this research advances automated forest fire detection systems.