Qvac

Qvac is a local AI SDK and set of libraries designed to enable developers to build private, cross-platform, peer-to-peer AI applications. It supports running large language models (LLMs), speech-to-text, translation, and other AI tasks entirely on-device across Linux, macOS, Windows, Android, and iOS. The tool is aimed at developers who prioritize data privacy, offline operation, and edge computing without relying on cloud services. For whom: Developers and teams building AI-powered apps that require local processing, such as privacy-focused chatbots, offline translators, or voice assistants. It is particularly suited for those who need to deploy on multiple operating systems and mobile devices. How it works: Qvac provides a unified SDK that abstracts the underlying local models (like Llama, Whisper, etc.) and handles inference on the user's device. The architecture is peer-to-peer, meaning data stays on-device or is shared directly between devices without a central server. It includes libraries for various languages and platforms, allowing integration into existing projects. What makes it different: Unlike most AI solutions that require internet connectivity and cloud APIs, Qvac runs everything locally, ensuring data never leaves the device. Its cross-platform support (including mobile OS) and peer-to-peer capabilities set it apart from other local AI runners that typically target only desktop or server environments.