☼ Good morning. Run the morning protocol?
Settings
Plays a thinking sound while model responds
Volume
Demo mode
Hide memories & history for demos
Thinking
Show model's reasoning
Speak Thinking
Read thinking aloud via TTS
Debug
Show response timing metrics
Auto-listen
Auto-send after silence
Interrupt by speaking
Speak over Jarvis to interrupt
Headphone mode
No echo cancellation needed — faster interrupts
Thinking sounds
Hmm, okay, so... while thinking
Online STT (Speechmatics)
Cloud STT with speaker diarization
Voice enrollment
Not enrolled
Enroll your voice to enable speaker-verified interruption — filters TV and background noise
Garden Mode 🌹
Dreamy roses in the background
Changes voice and response language
Hint for speech recognition (auto-detect recommended)
Runs systemctl restart jarvis. Page will reconnect automatically.
Reset audio pipeline without refreshing.
Force-reset audio output for headphones (iOS).
Debug Panel
On-screen logging for troubleshooting