Tether, the company most people associate with the world’s largest stablecoin, just released an open-source brain-to-text engine that runs entirely on your device. No cloud. No data leaving your hardware. No third party reading your neural signals.
Tether has been quietly building out its AI division, Tether EVO, and the latest release slots into its broader QVAC framework, a privacy-first toolkit designed to keep sensitive data exactly where it belongs: with the user.
What Tether actually built
The project is called BrainWhisperer, and it does what the name suggests. It decodes brain signals into text, running entirely on local hardware through the QuantumVerse Automatic Computer (QVAC) ecosystem.
QVAC launched on April 9, 2026, as a cross-platform, fully open-source software development kit for on-device AI inference and application development.
The brain-to-text component leverages OpenAI’s Whisper architecture, specifically through whisper.cpp, a lightweight implementation that handles speech-related tasks. But instead of transcribing audio from a microphone, BrainWhisperer applies similar decoding principles to neural signals.
Tether EVO placed fourth out of 466 participants in the Brain-to-Text ’25 Kaggle Competition on February 12, 2026. The team achieved a word error rate of just 1.78%. For context, that means roughly 98 out of every 100 words were decoded correctly from brain signals.
Why privacy matters more here than anywhere else
QVAC keeps everything local. CEO Paolo Ardoino has emphasized that full control of neural data remains with the user, a design philosophy baked into the architecture rather than bolted on as an afterthought.
The open-source component adds another layer of accountability. When the code is public, anyone can audit it, verify that data isn’t being exfiltrated, and fork the project to build their own implementation.
Tether’s bigger play: the machine economy
The QVAC framework is designed to support what the company calls a “Machine Economy,” a network of autonomous agents that can transact in Bitcoin and USDT across decentralized applications.
The QVAC SDK extends beyond brain-computer interfaces to support robotic toolkit applications, suggesting Tether EVO envisions a world where local AI doesn’t just decode your thoughts but also controls physical systems and manages financial transactions autonomously.
What this means for investors and the crypto market
For developers, the open-source QVAC SDK represents an opportunity to build applications that combine AI inference with crypto-native payment rails without relying on centralized cloud infrastructure. The competitive showing at the Kaggle competition serves as a credibility signal that the underlying technology actually works at a high level.
The QVAC GitHub presence continues to evolve, suggesting active development rather than a one-time announcement.