Loading...
Found 25 Skills
Fully offline speech-to-text via the Vosk library — streaming recognition, 16 kHz PCM, no network required after model download.