Tether CEO Paolo Ardoino has unveiled a new AI assistant named QVAC, demonstrating its ability to run complex tasks entirely on a local device without relying on cloud servers. In a demo shared on social media platform X, QVAC used natural language commands to create and assign tasks in the project management tool Asana. The system was notably running on a laptop with a "below average" GPU, highlighting its efficiency and challenging the industry norm of requiring powerful, centralized cloud infrastructure.
The core technology behind QVAC is its use of "100% local inference and reasoning" and support for the Model Context Protocol (MCP). This protocol allows the AI to connect to various external tools and services, enabling it to add new "skills" without altering its core model. Ardoino confirmed that Tether plans to open-source the QVAC assistant soon, inviting developers to build upon it and accelerating its integration into various applications.
This announcement marks a significant strategic expansion for Tether, the company best known for issuing the USDT stablecoin. It aligns with Tether's broader push into decentralized technologies, including previous hints about local AI applications, data tools, and robotics software. The primary selling point of QVAC is enhanced data privacy and user control, as sensitive information never leaves the user's device—a feature of particular interest in finance and crypto sectors.
Industry experts note the potential for QVAC to democratize access to powerful AI by making it viable on consumer-grade hardware, which could foster innovation in low-connectivity regions and reduce dependence on major cloud AI providers. The move positions Tether at the intersection of the AI and blockchain revolutions, with future potential for synergy where local AI agents could privately manage cryptocurrency portfolios or execute DeFi strategies.
