tetherto/qvac

QVAC - Local AI SDK and libraries for building private, cross-platform, peer-to-peer AI applications. Run LLMs, speech-to-text, translation, and more locally on Linux, macOS, Windows, Android, and iOS.

What it solves

QVAC addresses the privacy and dependency concerns of cloud-based AI by enabling local-first execution. It allows developers to build AI systems that run directly on a user's machine—across Linux, macOS, Windows, Android, and iOS—without relying on third-party APIs or SaaS providers. It also enables peer-to-peer (P2P) inference, allowing users to delegate AI tasks to other peers in a decentralized network.

How it works

The project provides a unified JavaScript SDK that serves as the primary entry point for developers. This SDK allows for the loading of models and execution of various tasks. It can function as a local model provider via an HTTP server that is compatible with the OpenAI API, making it easy to integrate into existing AI tools. The ecosystem supports a wide range of tasks including text generation, image/video generation, speech transcription, and even vision-language-action for robotics.

Who it’s for

  • Developers looking to build privacy-focused, local-first AI applications.
  • System architects building decentralized, unstoppable AI networks using P2P capabilities.
  • Users who want to use their own hardware as a local model provider for existing AI tools.

Highlights

  • Local-first execution for LLMs, speech, and RAG without cloud dependency
  • Peer-to-peer (P2P) capabilities for decentralized AI inference
  • Cross-platform support including Node.js, Bare runtime, and Expo
  • OpenAI-compatible API for easy integration with existing ecosystems
  • Broad AI task support ranging from text and image generation to BCI and robotics control