Optimized for edge hardware, this open-source tool provides fast, private on-device speech recognition for real-time transcription, command and speaker identification, lower word-error rates than Whisper models, and multi-language support over varied platforms.
United States


































































