This project uses machine learning to handle hCaptcha without reliance on browser plugins or third-party anti-captcha services. It leverages ResNet and YOLOv8 to address classification and detection issues, employing ONNX to package the model, thereby reducing the barrier to entry for users. It allows the code to run on devices without GPUs, and eliminates the need to install dependencies like PyTorch, enabling an elegant approach to dealing with hCaptcha.