Mountain landscape

Settings Reference

Open Settings from the Settings tab in the sidebar. Settings are organized into five sub-tabs: General, Dictation, Corrections, Models, and About.

General

SettingTypeDefaultDescription
ThemeSystem / Light / DarkSystemApp color scheme. System follows your macOS appearance.
Hide from app switcherToggleOffRemove Tota from the Cmd+Tab app switcher. The menu bar icon remains accessible.
Monitor clipboard historyToggleOffRecord text copied to the system clipboard.
Archive transcriptionsToggleOnAuto-archive transcriptions older than 3 days as Markdown files. Includes a folder picker and an Open in Finder button.
Replay the welcome tourButtonRevisit the first-run setup and feature walkthrough.

Dictation

Permissions

Status rows for Accessibility, Microphone, and Speech Recognition. Each shows a green "Granted" check or a button to grant access via System Settings.

Audio Input

SettingTypeDefaultDescription
DevicePickerAutomatic (Built-in Mic)Select which microphone to use. If your preferred mic disconnects, Tota falls back to another device and switches back automatically when it reconnects.

Keyboard Shortcuts

SettingTypeDefaultDescription
Toggle RecordingShortcut recorderCmd+Shift+RGlobal shortcut to start/stop recording.
Hold to SpeakPickerOptionModifier key to hold for recording. Options: Option, Control, Function (fn), or Disabled.
Learn CorrectionsShortcut recorderCmd+Shift+DGlobal shortcut to manually commit dictionary edits.
Reset to DefaultsButtonRestore all shortcuts to their factory defaults.

Behavior

SettingTypeDefaultDescription
Auto-paste after transcriptionToggleOnAutomatically insert transcribed text at the cursor.
"Hey Tota" wake wordToggleOffListen in the background and start recording when you say "Hey Tota", "wake up", or "hey Mac". Requires Speech Recognition permission.
Auto-stop recording on silenceToggleOffAutomatically stop recording after a pause. Silence duration adjustable from 0.5 to 10 seconds (default 2).

Transcription

SettingTypeDefaultDescription
ModeTranscribe / TranslateTranscribeTranslate types the English translation of whatever language you speak. Requires a multilingual model (Large v3 or Large v3 Turbo).
LanguagePickerEnglishWith a multilingual model: Auto-detect or one of 13 languages.
English SpellingUS / UKUSUK keeps British spellings (colour, realise, centre) in transcriptions and Smart Formatting.

Corrections

SettingTypeDefaultDescription
Smart correctionsToggleOnMaster switch: apply your dictionary and custom word fixes to each transcription. Turn off to keep raw transcription.
Automatically learn correctionsToggleOnWatch for edits after pasting and learn word corrections. When off, use the Learn Corrections shortcut instead.
Phonetic vocabulary matchingToggleOnFix mishearings of your active glossary's terms (e.g. "versel" → "Vercel").
Correction thresholdStepper (1–10)2Occurrences required before a word swap is auto-applied. Word edits (e.g. recieve → receive) apply immediately.
Auto-pick glossary by appToggleOffChoose the best glossary for the app you're dictating into (e.g. Slack → Names, VS Code → Development).

Models

Whisper Model

Radio list of transcription models with a Load Model button. Base (English) is built-in; other models download from HuggingFace on first use. Selecting a model does not switch until you click Load Model — a notice reminds you if dictation is still using the previous model.

Smart Formatting

Enable Smart Formatting (default: off) turns on the on-device AI cleanup pass. First-time use downloads the formatting model (~2.3 GB), with progress shown inline. See the Smart Formatting article for details.

About

License

Displays your current license status (Trial, Active, Expired), days remaining, and options to activate or deactivate a license key.

Updates

SettingTypeDescription
Check for UpdatesButtonManually check for a new version. Disabled if your license is expired or inactive.
Automatically check for updatesToggleEnable or disable automatic update checks.

Acknowledgements and Support

Buttons for third-party library credits and support resources.