Tether launches open-source QVAC SDK to support localized AI and cross-platform deployment

According to Odaily, stablecoin issuer Tether has announced the launch of QVAC SDK, a fully open-source, cross-platform software development kit designed to be a universal AI building block, enabling the operation, training, and evolution of artificial intelligence on any device and operating system. Functionally, this SDK supports a wide range of AI capabilities, including large language models (LLM), speech recognition, text generation, translation, OCR, and speech synthesis, all accessible through a unified API. Its backend is based on QVAC Fabric (originating from the llama.cpp branch), integrated with local inference engines such as whisper.cpp, Parakeet, and Bergamot.

QVAC SDK is reported to adopt a “local-first” architecture, allowing developers to build and run AI applications directly on end devices (such as smartphones, PCs, servers, and more), achieving unified code deployment for iOS, Android, Windows, macOS, and Linux without the need for platform-specific adaptation or rewriting. Additionally, QVAC SDK comes with built-in peer-to-peer (P2P) capabilities based on the Holepunch technology stack, supporting decentralized model distribution, inference, and, in the future, distributed training, reducing reliance on centralized services.

Tether stated that it will continue to increase investment to expand the QVAC open-source ecosystem, covering advanced fields such as robotics and brain-computer interfaces, and promote a shift in AI deployment from cloud-based to local models to meet requirements for privacy, low latency, and high reliability.