The company behind the world’s largest stablecoin is on a hiring spree, and it has nothing to do with dollar reserves. Tether is actively recruiting inference engineers for its QVAC initiative, a project designed to run AI models locally on consumer devices like smartphones and laptops, no cloud required.

The open roles span multiple seniority levels, including Lead AI Inference Engineer, Senior AI Inference Engineer, and AI Inference Engineer positions within the QVAC division. Candidates are expected to bring expertise in frameworks like llama.cpp and ggml, both widely used in the open-source community for running large language models on modest hardware. The focus is on optimizing Tether’s C++ runtime layer for edge devices.

QVAC: Tether’s bet on local-first AI

QVAC was initially unveiled in May 2025 as a decentralized platform for running AI agents directly on user hardware. The QVAC Fabric LLM now supports inference on consumer-grade devices across various GPU architectures, spanning smartphones to desktops.

The most tangible proof of progress came on May 7, 2026, with the launch of QVAC MedPsy. This is a medical language model built specifically for local execution on smartphones and wearables. Tether claims it delivers performance matching or exceeding traditional cloud-based models.

Four days after the MedPsy launch, on May 11, 2026, Tether announced a developer grants program aimed at bolstering local-first AI and payment infrastructure. The program is structured without payout caps, instead tying compensation directly to technical deliverables.

Why a stablecoin company cares about AI inference

Tether has been building out what it calls Tether Data, a division focused on peer-to-peer platforms and open-source AI tools. The thesis is that controlling the full stack, from the payment layer (USDT) to the communication layer to the AI layer, creates an ecosystem that doesn’t depend on big tech intermediaries.

CEO Paolo Ardoino has hinted at future developments that could further integrate these pieces, suggesting that on-device AI combined with decentralized payments could unlock entirely new user experiences.

What this means for investors

The competitive landscape here is interesting. Most crypto-AI projects are focused on decentralized compute networks, essentially building cloud alternatives using distributed GPU resources. Tether is going the opposite direction, eliminating the need for remote compute altogether. That means QVAC isn’t directly competing with projects like Render or Akash. It’s competing with Apple Intelligence and Google’s on-device AI efforts.

The grants program is worth monitoring closely. Uncapped, deliverable-based payouts tend to attract serious builders rather than grant farmers. If QVAC’s developer community grows meaningfully over the next two quarters, it would validate Tether’s thesis that there’s genuine demand for local-first AI tools built outside the big tech ecosystem.