Tether AI expanded its on-device AI toolkit with the release of QVAC SDK 0.16 on July 27, 2026. This update introduces new features including Ideogram 4 for image generation, LTX-2 enabling synchronized video with audio, and local support for NVIDIA’s GR00T robotics model.
According to Tether's CEO Paolo Ardoino, the company is pushing beyond its reputation for stablecoins by focusing on building the infrastructure that powers private AI directly on users' devices. This approach aligns with their concept of a “Stable Intelligence Era,” which extends their financial sovereignty principles to the AI space.
What the update offers developers
QVAC SDK 0.16 allows developers to generate complex images and produce videos with synchronized audio without relying on cloud processing. The integration of NVIDIA’s GR00T model is particularly notable, as it facilitates running advanced robotics AI locally. This means applications involving humanoid robots can process data safely without sending sensitive information to external servers.
The SDK is open-source under Apache 2.0 and compatible across multiple operating systems through Node.js. It supports multimodal AI models that handle text, vision, speech, and robotics, leveraging peer-to-peer capabilities through the Holepunch stack. Recent updates before this release included TurboQuant in June and other enhancements in July.
also the improved OpenAI-compatible server within QVAC simplifies the process of integrating AI agents into existing developer pipelines, removing the need for extensive rewiring.
While QVAC itself is not linked to any native cryptocurrency token, Tether’s existing USDT stablecoin remains the reference point for transactions in decentralized apps that might build on this toolkit.
