First launch and permissions

The tutorial window opens on first launch. Go through the eight steps and press Get started on the last screen.

Press Get started at the end

If you close the tutorial window with its close button, setup counts as unfinished. In that state dictated text is only copied to the clipboard instead of being inserted at the cursor, and the tutorial opens again on the next launch. Skipping individual steps with Skip is fine.

The steps at a glance

Step	Screen title	What to do
1	A menu bar app that turns speech into text	Read the introduction and press Next.
2	Grant permissions	Allow Microphone and Accessibility.
3	Download the speech model	Pick a model. Large v3 Turbo is the recommended default.
4	Groq API key (optional)	Only for cloud speech recognition. Usually leave it blank.
5	LLM correction (good to know)	An overview of correction engines and modes. You set this up later.
6	Where should the app appear	Choose the menu bar, the Dock, or both.
7	What else you can do	One screen with the features beyond dictation and their shortcuts. Nothing to set up.
8	You are set	Check the recording shortcut and readiness, then press Get started.

Granting permissions

Two permissions are required. The rest are only needed for specific features.

Permission	Why	Without it
Microphone	Recording your voice	Recording does not start.
Accessibility	Global shortcuts and pasting at the cursor	Shortcuts do nothing and text is not inserted.
System Audio (Screen Recording)	Recording sound played on your Mac, such as the other side of a call	Only the microphone is recorded.
Automation · Google Chrome / iTerm2	Returning to the original tab, field, or terminal session after recording	Sogon only activates the app and pastes.

1. On the permissions screen, press Grant next to Microphone and allow it in the macOS dialog.
2. Press Grant next to Accessibility. When System Settings → Privacy & Security → Accessibility opens, turn Sogon on.
3. Return to the tutorial and check that both items show Granted.

Sogon only asks for permissions when you press a button. Launching the app or starting a recording never opens a permission dialog. To check later, open Settings → General → Permissions.

Downloading a model

Picking a model in step 3 starts the download right away. You can press Next while it downloads, and closing the window doesn't stop it.

Model	Size	Choose it when
Large v3 Turbo compressed	about 632MB	You want the quickest start and to save disk space
Large v3 Turbo	about 1.6GB	You are not sure (app default)
Large v3	about 3.1GB	Accuracy matters more than speed

If you skip without picking a model, the default model is downloaded on your first recording. You can change models later in Settings → STT.

Reading the readiness list

The Readiness list on the last screen shows the microphone, the global shortcut, and the speech model. When all are green, Everything is ready appears.

If you see Permission is granted but the shortcut listener is not attached yet, press Get started, then quit and relaunch Sogon.

Seeing the tutorial again

Open it any time from Settings → General → Tutorial → Show again. It doesn't reset your current settings.

Next: Your first dictation