Settings
Audio devices
Select your preferred microphone input device from the dropdown in Settings. MeetingDebrief lists all available audio input devices on your system. If you connect a new device, it appears in the list automatically. The app follows your system default input device if it changes during a recording.
Toggle Include system audio to enable or disable capturing audio output from your computer during recordings. On Windows, System audio device lets you pick which output to capture, for example the headset your meeting app plays through. If a saved device disappears, MeetingDebrief tells you instead of quietly recording only your microphone.
Detect speakers labels transcript lines as Me or Other by comparing your microphone with system audio. Turn it off if lines are often mislabeled.
Use the Microphone test to verify your mic is working and see volume levels before starting a recording.
Model management
The models section displays all available transcription and summarization models. For each model you can:
- Download: download a model you have not yet installed.
- Select: set the model as your active transcription or summarization model.
- Delete: remove a downloaded model to free disk space.
Switching models does not affect existing transcripts or summaries.
Once Parakeet is installed, Keep the model loaded sets how long it stays in memory between uses. See Transcription.
The Summary language dropdown lets you force summaries into a specific language instead of relying on auto-detection from the transcript.
GPU acceleration
On macOS with Apple Silicon, GPU acceleration via Metal is used automatically for both transcription and summarization.
On Windows with an NVIDIA CUDA-compatible GPU, you can download the CUDA worker from the models section to enable GPU-accelerated transcription and summarization. This can make processing up to 10x faster. The GPU status is displayed in Settings with options to re-enable if it was previously unavailable.
Vocabulary
Add the names, acronyms, and jargon MeetingDebrief should spell correctly. See Vocabulary.
Dictation
Turn dictation on or off, set its shortcut (default Control+Alt+D), choose how it starts and stops, whether it pastes for you, sounds, muting other audio, where the dictation window appears, and how long dictation audio is kept. See Dictation.
Appearance
In General → Appearance, choose System to follow your operating system, or pin MeetingDebrief to Light or Dark. Every window follows the choice, including the floating recorder, the dictation window, setup, and updates.
Keyboard shortcuts
Customize the global keyboard shortcut used to toggle the overlay window. The default shortcut is CmdOrCtrl+Shift+R. Click the shortcut field and press your preferred key combination to change it.
Automation
Two automation options control what happens after a recording stops:
- Auto-transcribe: automatically start transcription when a recording finishes.
- Auto-summarize: automatically generate a summary after transcription completes, using your default template.
Both options are independent. You can enable auto-transcribe without auto-summarize, or both together.
Preferences
- Keep audio files: when disabled, MeetingDebrief deletes the audio recording after transcription completes. Enable this to retain audio files alongside transcripts. When enabled, you can set a retention period in days (0 keeps files forever).
- Live transcription during recording: displays a real-time transcript while you record. May be less accurate than the final transcription.
- Performance priority: controls how much system resources MeetingDebrief uses during processing. Options are Full (fastest results), Balanced (shares resources with other apps), and Background (minimal impact on other apps).
Dates and times follow your OS locale settings automatically.
Re-run setup wizard
Click this button to restart the initial setup process. Useful for reconfiguring permissions or re-downloading models. Your existing sessions and data are not affected.