Modes

A mode decides what happens to your words after YapToText turns them into text. Seven modes ship with the app: Auto, Raw Transcription, Clean Up, Note, Email, Message, and Code Comment. Raw Transcription types exactly what you said, with punctuation, and never touches the cleanup model. The rest hand your transcript to the cleanup model with written instructions, each shaping the same words for a different kind of writing.

You pick a mode from the menu bar popover, or by pressing a number key from 1 to 9 while you are dictating. Every mode is editable, you can write your own, and any app can have its own default mode.

What each built-in mode does

• Auto reads each dictation and picks the right mode by itself. A letter with a greeting and sign-off comes out as an email, casual chat stays as spoken, and everything else gets the standard cleanup. It has its own page: Auto mode.
• Raw Transcription is fast, literal speech-to-text with basic punctuation. No cleanup model runs.
• Clean Up removes filler words and false starts and fixes grammar, capitalization, and punctuation. It also fixes words the microphone clearly misheard, picking the word you most likely meant from the words around it. It is the mode I recommend as a default. It is conservative: it never deletes a sentence, never shortens or summarizes, and never swaps text for an ellipsis. Casual words stay casual and blunt words stay blunt. Spoken numbers become digits, but it will not add a unit or a time of day you did not say. It does not translate.
• Note turns spoken ideas into a structured note with short headings and bullet points, and pulls action items into a list. Bullets and headings can only restate what you said. It never adds items, actions, or reminders you did not speak.
• Email formats your dictation as a clear email body while keeping your voice. It never writes a Subject line. It adds a greeting only if you named or clearly implied a recipient, and if you end with your own name, that becomes the sign-off on its own line. It never invents pleasantries or placeholder text like [Your Name].
• Message lightly cleans a chat message. Filler words go, obvious grammar gets fixed, and the tone stays casual.
• Code Comment is for developers. It rewrites your dictation as concise, imperative text for a code comment or a commit message. Identifiers, casing, and technical terms stay exactly as spoken, with no quotes or code fences.

Clean Up, Note, Email, and Message all keep the language you spoke and add nothing you did not say. Clean Up, Note, and Message also never answer a question that happens to be inside your text. Both models run on your Mac: out of the box, Whisper Large v3 Turbo hears you and the bundled cleanup model rewrites. Nothing is uploaded.

Switching modes with the number keys

Press 1 through 9 while dictating to pick the mode for that dictation. The order matches your Modes list, so 1 is always the first mode in the list. Auto sits first, which is why Raw Transcription shows as 2 above.

The number keys are only captured while you are actually dictating, so they never interfere with normal typing. Every key is listed on Keys and shortcuts.

Editing a mode, or making your own

Every mode is editable, including the built-in ones. Open a mode, change its instructions, and the cleanup model follows them. If Email keeps adding a greeting you do not want, or Note uses more headings than you like, rewrite the instructions until it does what you want. They are plain sentences.

You can also create entirely new modes. The quickest way is to duplicate one that is close and edit the copy. A mode can carry its own choices: whether the cleanup model runs at all, and which speech and cleanup models it uses. A mode that carries its own model choice always wins over the Energy defaults.

A default mode for each app

Any app can have its own default mode, so dictation in that app uses that mode without you picking it. Put Mail on Email, Messages on Message, and your editor on Code Comment, and you stop thinking about modes in those apps.

Running a mode on text you already have

Modes are not only for live dictation. Select text in any app, then open the Regenerate menu in the menu bar popover and pick a mode. The rewrite replaces the selection in place. Reading the selection needs the Accessibility permission. The same menu regenerates your last dictation as any mode, and History can regenerate any old dictation with a different mode, since every entry keeps the raw transcript, the final text, and which mode ran. Utility does the same inside the app: paste anything and run it through a mode.

Questions

All help topics · YapToText home