Get started with Soufflé on Mac

Start with a supported Mac, install the app, then download one speech model. Dictation needs no account or AI provider. Add meeting audio and summaries when you need them.

Updated

Start here

  1. Install Soufflé with Homebrew or the signed DMG from GitHub, then open it from Applications.
  2. Grant Microphone access, choose a speech model and wait for its first download. Keep the default Kyutai 1B model for streaming French and English.
  3. Record a short dictation and check the text. Enable auto-paste and grant Accessibility access if you want the text inserted into the app you were using.

Install

With Homebrew:

brew install --cask damione1/tap/souffle

Or download the .dmg from the latest release and drag Soufflé into Applications. Every build is signed and notarized, so it opens normally: no right-click detour, no Gatekeeper warning.

On first launch the setup wizard downloads a speech model, between 670 MB and 5.6 GB depending on which one you pick. That download is the only moment Soufflé needs the network in order to transcribe.

Requirements

MacApple Silicon. There is no Intel build.
macOS 13+Dictation, and meetings recorded from your microphone.
macOS 14.4+Meetings that also capture the other participants. System-audio capture uses the Core Audio process tap, which does not exist before 14.4. On macOS 13 a meeting still records, and the transcript is labelled Mic only.
Disk670 MB to 5.6 GB for the speech model, downloaded on first use.
SummariesApple Intelligence, which needs macOS 26 or newer, or a local Ollama on any supported version. Transcription needs neither.

Permissions

The setup wizard has an optional Permissions step; skip it and Soufflé asks for each of these as you use the feature that needs it. Notifications are the exception, requested when the app launches. Settings → System → Permissions shows the current state of each and links straight to the matching System Settings pane.

MicrophoneRecords your voice so it can be transcribed. Needed for everything.
System AudioCaptures what the other participants say, without a virtual audio device. Meetings on macOS 14.4+.
AccessibilityInserts the text into the app you were using, reads back your corrections when “learn from edits” is on, and installs the native event tap that lets a dictation shortcut bind to a single key (a lone modifier, or F5–F12).
CalendarReads today's events to list them on Home and notify you when one starts. Optional, and off until you turn it on.
NotificationsCalendar reminders, a meeting about to stop on its own, a paste that failed. Requested at launch; refusing it only silences those alerts.

Refusing one disables only the feature that needs it. Without System Audio a meeting still records your microphone and the transcript is labelled Mic only. Without Accessibility the dictated text is copied to the clipboard instead of being inserted for you, and a single-key shortcut cannot install its event tap; classic modifier shortcuts still work.

First run

A setup wizard covers the permissions Soufflé needs, which microphone to use, the speech model to download, and the dictation shortcut. The default shortcut is ⌘⇧Space.

The shortcut step also offers auto-paste, off until you tick it: when it is on and you stop dictating, the text goes straight into the app you were in. Left off, the text is put on the clipboard instead. A fresh install finishes with Launch at login on; change that later in Settings → System. Everything here can be changed later in Settings.

Settings reference

Six tabs. The settings worth knowing about:

Insertion method
InterfaceClipboard and ⌘V, simulated typing, or a direct Accessibility write. Change it when an app refuses the paste.
Auto-paste
InterfaceOn: the text goes into the app you were using. Off, the default: it is copied to the clipboard.
Paste delay
InterfaceHow long to wait before pasting. Raise it if the text lands in the wrong app.
Language and theme
InterfaceThe interface in French or English, and a light, dark or system appearance.
Dictation sounds
InterfaceStart and stop cues, with a volume.
Hide recording overlay
InterfaceHide the floating HUD while dictating or in a meeting. Drag it when visible to move it; the position is remembered. A second switch, on by default, keeps it out of screen captures.
Dictation polish
AIThe local LLM pass and its prompt templates, which you can edit. On by default when Apple Intelligence or Ollama is available.
Custom dictionary
TranscriptionProper nouns and technical terms to spell correctly, filled in by hand or learned from your corrections.
Voice snippets
TranscriptionTrigger phrases that expand to a prepared block of text at the start of a dictation. Matching skips the polish pass.
Microphone priority
AudioAn ordered list of preferred inputs, a separate choice for clamshell mode, and a switch that keeps Bluetooth headsets from being used as microphones so they stay in stereo.
Meeting auto-stop
AudioStop after 5, 10, 15 or 30 minutes without speech, with a hard failsafe behind it.
Meeting transcription language
AudioA hint for language detection in meetings, never forced on the engine.
Model unload
TranscriptionFree the model's RAM after 5, 15 or 60 idle minutes, or never.
Calendar
MeetingsOff until you turn it on. Lists today's events from the calendars you keep ticked, reminds you a few minutes before one (2 by default), and notifies you when one begins while a call is using the microphone.
Summarization
AIProvider, Ollama server and model, and the summary templates.
Launch at login
SystemOpen Soufflé automatically when you start your Mac. On by default after a fresh setup wizard.
Meeting audio
SystemOff, keep 7 days, keep 30 days, or keep until you delete it.
Model storage
SystemDeletes the selected speech model from disk, once it is downloaded and not loaded.
Update check
SystemAsks GitHub once a day whether a newer release exists. Turn it off in About. An update is downloaded only when you press Download update, and installed with Install and restart, which asks you to stop a running meeting first.