Show HN: Free On-Device AI SDK to Run PyTorch on Mobile NPUs (Open Source) (github.com)

ZETIC.ai has announced the release of MLange, an open-source SDK designed to facilitate the deployment of AI applications on mobile devices utilizing Neural Processing Units (NPUs). This innovative platform allows developers to transform their AI models into production-ready mobile apps within hours— a process that previously required extensive hardware-specific tuning. MLange supports real-time, zero-latency performance while ensuring user privacy by running all operations locally, eliminating dependencies on internet connectivity or cloud services. Significantly, MLange accelerates AI application development by providing automated model optimization and a unified API for both Android and iOS. With just three lines of code needed to integrate and a library of various pre-optimized models available for use (including applications in computer vision, NLP, and audio processing), developers can quickly prototype and deploy robust AI solutions. The platform has been benchmarked on over 200 devices, showcasing its reliability and versatility. This democratizes access to advanced AI capabilities, empowering developers to create diverse applications while maintaining data security and operational efficiency.