Command line

Drive the service from scripts, launchers and any hotkey system.

Recording

# Start recording
hyprwhspr record start

# Start recording with specific language
hyprwhspr record start --lang it    # Italian
hyprwhspr record start --lang de    # German
hyprwhspr record start --lang es    # Spanish

# Stop recording (transcribes and pastes)
hyprwhspr record stop

# Push-to-talk key-up: stop, or latch a long hold
hyprwhspr record release

# Cancel recording (discards audio, no transcription)
hyprwhspr record cancel

# Toggle recording on/off
hyprwhspr record toggle
hyprwhspr record toggle --lang it   # Toggle with language override

# Check current status
hyprwhspr record status

# Capture: trigger a recording and stream the transcription to stdout
# Blocks until transcription is complete. Suppresses text injection — use for scripting/piping.
# Self-triggers a recording if none is in progress; attaches to an in-flight recording if one is.
hyprwhspr record capture
hyprwhspr record capture --lang it   # Capture with language override

# Diagnose how the running daemon would process a captured transcript
hyprwhspr record capture --trace-processing

Trace processing

--trace-processing prints one UTF-8 JSON document using the running daemon’s settings. It reports exact raw and preprocessed text, backend/model, recording and silence settings, symbol/hook state, vad_mode, and boundary_mode. It does not run hooks, append space, save the transcript, or inject text.

vad_mode: none; silero_filter (pre-inference filtering); silero_segmented (ONNX segmentation after the reported duration gate); server_vad; manual_commit; or provider_managed. boundary_mode is manual_stop, silence_auto_stop, or continuous_silence. Continuous silence flushing and realtime server VAD can turn pauses into independently punctuated segments. Trace does not change VAD or silence defaults.

File transcription

Files in. Words out. WAV and MP3; local and REST backends.

hyprwhspr transcribe recording.mp3
hyprwhspr transcribe recording.wav -o transcript.txt
hyprwhspr transcribe recording.wav --lang fr --clean

Stdout by default; -o writes UTF-8 text. --lang sets the language. --clean applies configured cleanup, without pasting or running hooks.

An idle service lends its loaded model. A busy service says no. No service: the command loads the backend itself.

File transcription does not work with the realtime WebSocket backend.