Tether CEO Paolo Ardoino revealed a first public demo of “QVAC,” an artificial intelligence assistant currently under development by Tether.
The preview suggests the company is expanding beyond stablecoins into AI infrastructure, with a strong focus on local execution and privacy-preserving design.
According to Ardoino, QVAC supports multiple skills through the Model Context Protocol (MCP) and is capable of running 100% local inference and reasoning, even on a sub-average laptop GPU.
Testing Tether's QVAC AI Assistant
many skills already supported via MCP, using Asana in the example below (on a sub-average laptop GPU)
100% local inference/reasoning
soon opensource pic.twitter.com/sYi91QhjVC— Paolo Ardoino 🤖 (@paoloardoino) February 12, 2026
100% Local AI Processing
One of the most notable aspects of QVAC is its fully local architecture. Ardoino emphasized that both inference and reasoning occur directly on the user’s device, rather than relying on remote cloud servers.
In the shared demo, the assistant interacted with Asana to create tasks using MCP-enabled tool calls, demonstrating practical workflow integration. The interface displayed structured task creation, including titles, IDs, and assigned users, suggesting enterprise productivity use cases.
Running locally offers several potential advantages:
- Greater data privacy
- Reduced dependency on centralized AI infrastructure
- Lower latency for task execution
If validated at scale, this approach would differentiate QVAC from most mainstream AI assistants that depend heavily on cloud-based computation.
MCP Skill Integration and Open-Source Plans
QVAC leverages the Model Context Protocol to enable modular “skills,” allowing it to integrate with external tools and services. In the example shared, the assistant created an Asana task and subtask autonomously, highlighting early-stage automation capabilities.
Ardoino also confirmed that the project will be open-sourced soon, signaling Tether’s intention to involve the broader developer community.
Strategic Context
While Tether is primarily known as the issuer of USDT, the development of QVAC suggests a broader strategic expansion into AI tooling and decentralized infrastructure. By prioritizing local computation and open-source collaboration, Tether appears to be positioning itself at the intersection of AI sovereignty and digital asset ecosystems.
Further technical details, release timelines, and architecture specifics are expected as the project moves toward open-source deployment.
