watadarkstar/react-native-nsfw-detector

This is a library for mobile app developers building iOS apps with React Native or Expo. It checks whether an image contains nudity or other unsafe visual content, and it does that check entirely on the device, with no internet connection required. The underlying technology is Apple's CoreML, which lets apps run trained AI models locally on an iPhone or iPad. The way it works is straightforward: you pass the library a path to an image, and it returns a confidence score between 0 and 1. A score above 0.9 means the image is likely unsafe. A score below that suggests it is safe to display. You call one function, get one number back, and decide what to do with it. There are no accounts, no API keys, and no data sent to any server. The library requires iOS 13 or newer, React Native version 0.70 or above, and a real iPhone or iPad for accurate results. The README warns clearly that running on the iOS Simulator produces unreliable scores, so testing on a physical device is necessary. Xcode 10 or above is also required because the built-in AI model was created with Apple's CreateML tool. Setup follows the standard pattern for React Native packages: install via npm or yarn, then run your app. The repository includes a working example Expo app in the example directory if you want to see a full integration before writing your own. The project is MIT licensed and credits the original NSFWDetector library by the LOVOO organization as the foundation for the detection model.