NudeNet provides Python and JavaScript libraries for detecting nudity and NSFW content in images using neural network models. The project offers pre-trained ONNX and PyTorch models (320n and 640m variants) built on ultralytics YOLOv8 architecture at different inference resolutions. Images can be processed as file paths, OpenCV arrays, or raw bytes, with support for both single-image and batch detection modes. The models classify detected regions into anatomical categories like BELLY_EXPOSED, with confidence scores for each detection.