Tether launches QVAC SDK for local AI across phones and desktops
The open source toolkit lets developers build and run AI on phones, laptops, and servers as Tether expands deeper into privacy focused infrastructure.
Share
Add us on Google by Estefano Gomez Apr. 9, 2026Tether has launched QVAC SDK, an open-source software development kit designed to let developers build, run, and fine-tune AI directly on local devices across major operating systems.
The company said applications built with the toolkit can run across iOS, Android, Windows, macOS, and Linux from a single codebase, without relying on cloud-only infrastructure.
Tether said QVAC SDK is designed to keep AI functions such as writing assistance, transcription, translation, image generation, search, and summarization running directly on consumer hardware, including smartphones, laptops, desktops, and servers. The company said that should improve privacy, lower latency, and keep services functioning even when connectivity is weak or remote servers go offline.
At the core of the stack is QVAC Fabric, which Tether describes as a fork of llama.cpp that supports text generation, embeddings, and multimodal workloads. The SDK also integrates whisper.cpp and Parakeet for speech-to-text and Bergamot for on-device translation, exposing those engines through a unified interface so developers can swap or combine capabilities without rewriting application logic.
Tether is also pitching QVAC as more than a standalone SDK. The company’s QVAC site describes the project as its answer to centralized AI, with a focus on private, local, and permissionless intelligence on any device. Earlier efforts tied to the QVAC initiative included QVAC Genesis II, a large synthetic educational dataset released in late 2025, and QVAC Health, a private AI health data hub.
Tether was in talks to lead a €1 billion funding round in Neura Robotics, underscoring Paolo Ardoino’s push to position the firm around a future shaped by autonomous machines and AI agents.
Disclosure: This article was edited by Estefano Gomez. For more information on how we create and review content, see our Editorial Policy.
