Moju V2 Realtime Web Console
Browser mic -> /voice/ws -> OpenAI Realtime (server_vad) + Sarvam TTS -> streamed audio reply.
One click starts server VAD mode: continuous mic streaming, server detects speech/silence, and speaking during playback triggers barge-in.
Current: -
WS
disconnected
Turn
idle
Turn ID
-
Voice Mode
server_vad
Audio Format
pcm16le
Mic RMS
0
VAD State
idle
Local stop->first audio
-
Local stop->total
-
Server stop->first audio
-
Server stop->total
-
Server STT / LLM
-
Server total
-
Bytes out / in
0 / 0
Frames out / chunks in
0 / 0
Partial Transcript
Final Transcript
Database Status
Redis
-
Supabase
-
Memory Pipeline
-
Shadow Mode
-
Memory
Identity Facts
-
Preference Facts
-
Context Facts
-
Pending Confirmations
none
Redis Hits
-
Redis Misses
-
Fallback Reads
-
Fallback Writes
-
Events