Whisper speech-to-text inference snap for Myna
Local, offline speech-to-text service exposing the Myna session API on a
Unix domain socket. Wraps Whisper models served via faster-whisper
(CTranslate2), with hardware-based engine selection and switchable model
sizes via the modelctl CLI.
The service receives PCM audio pushed by clients; it has no access to the
microphone or any audio hardware.