Tether AI Research Launches Multilingual Automatic Speech Recognition Model QVAC TranscribePsy

Foresight News reports that Tether AI Research has launched a multilingual automatic speech recognition model called QVAC TranscribePsy. The model features 254 million parameters and is designed to operate entirely on-device. Tether stated that TranscribePsy provides developers with an offline speech recognition solution, eliminating the need to send user audio to cloud servers. The model can process audio from smart glasses connected via Bluetooth or Wi-Fi in near real-time, and it can also run directly on the onboard processor of humanoid robots. TranscribePsy is specifically designed for rapid response, enabling low-latency natural speech interactions while keeping speech processing on-device.