NSFW JS is a JavaScript library developed by Unbeaning, aimed at identifying potentially inappropriate images directly on the client's browser. Powered by TensorFlowJS, an open-source machine learning library, NSFW JS boasts an impressive 93% accuracy rate in recognizing specific image patterns. Additionally, it incorporates CameraBlur Protection to automatically blur flagged images. Constantly evolving, this library offers regular updates and new models, all under the MIT license.

NSFW JS is a versatile tool that allows for content filtering, client-side image detection, and in-browser image evaluation without the need for server interaction. Lightweight at 4.2MB, it includes a mobile demo for on-the-go testing and encourages community contributions for further development.