VAD Tier 1 Edge

Zero-Latency Neural Voice Activity Detection, Word Timestamps & In-Browser Dictation

Mic: Off
VAD: Idle
ASR: Standby
PTT: [Space]
FSM: IDLE COMPUTE: WASM / CPU QUEUE: 0 INFER LATENCY: 0 ms CALIBRATED TRIGGER: 0.40
Initializing Whisper Model... 0%
Signal Presence Probability SILERO VAD ONNX
Confidence Score:
0.00
Thresholds & Calibration
ParameterDefaultLive ValueAdjust Range
Positive Speech Trigger 0.40 0.40
Negative Silence Cutoff 0.20 0.20
Soft-Knee Pre-Gain 1.5x 1.5x
Hold / Redemption Frames 8 8
Whisper Model Tier tiny.en
Synchronized Transcript Stream Click sentence to play audio
Real-Time Telemetry Stream 0 events