Mac Dictation Commands: The Complete Reference Guide

You're halfway through an email when the cursor stops keeping up with your thoughts. You press the dictation shortcut, speak naturally, and then discover that the Mac has treated a formatting instruction as ordinary text, missed a name, or mangled a technical string. The problem usually isn't that voice input has failed. It's that Mac dictation commands operate as a structured system with different modes, activation states, and app-dependent behavior.

Apple's built-in tools now cover more than speech-to-text. macOS includes commands for sentence structure, spacing, capitalization, navigation, and text control, while Voice Control separates natural dictation from character-by-character spelling and action-only commands. Apple's Mac dictation command documentation lists examples such as “new line,” “new paragraph,” “tab key,” “no space on/off,” “caps on/off,” “all caps,” “numeral,” and “roman numeral.” Once you understand which state the Mac is in, dictation becomes easier to predict, troubleshoot, and extend.

Table of Contents

• How Mac Dictation Starts and Signals Readiness
  • Read the interface as a state signal
• The Three Voice Modes You Need to Know
• Built-In Formatting and Navigation Commands
  • Punctuation and capitalization
  • Spacing and structure
  • Navigation and correction
• Spelling Mode and Phonetic Alpha Bravo Charlie Entry
• Stock Dictation Compared to AIDictation
• Custom Dictionary and App-Specific Context Rules
  • Separate vocabulary from context
• Keyboard Shortcuts and the Fn Twice Trigger
• App-Specific Commands Across Mail, Notes, Pages, and Editors
• Example Utterances for Clinical, Developer, and Multilingual Workflows
  • Clinical documentation
  • Developer notes and comments
  • Multilingual writing
• Why Enhanced Dictation and Custom Dictation Commands No Longer Apply
• Troubleshooting the Most Common Dictation Failures
• Quick-Reference Card for Mac Dictation Commands
  • Activation
  • Structure and punctuation
  • Capitalization and exact entry
  • Editing and app context

How Mac Dictation Starts and Signals Readiness

Dictation begins with a handoff between your keyboard, the active text field, and the speech engine. First, click where you want text to appear. Then press the Microphone key, use your configured keyboard shortcut, or choose Edit > Start Dictation, all activation paths Apple documents in its Mac dictation workflow.

Don't speak immediately after pressing the shortcut. Wait for the insertion point or microphone indicator to pulse, and listen for the system signal that confirms dictation is ready. The ready state matters because words spoken during the transition can be missed, inserted in the wrong field, or interpreted inconsistently. Apple's documentation describes this activation-and-ready sequence, including the configurable shortcut under Keyboard > Dictation.

Read the interface as a state signal

Once the indicator appears, speak in short, intentional units. For example, say:

“The report is ready period new paragraph Please review the summary.”

In a standard dictation field, the first sentence becomes text and “period” becomes punctuation. “New paragraph” changes the document structure. If you issue a navigation or editing instruction, the result depends on whether Voice Control Command mode is active and whether the current app supports that action.

A useful mental model is simple:

1. Idle: The Mac is waiting for activation.
2. Listening: The microphone indicator signals that speech can be captured.
3. Processing: Spoken words are converted into text or interpreted as commands.
4. Editing: You review the result, correct errors, or issue another action.

If nothing appears, check the insertion point before changing other settings. Dictation needs a usable text field, and the active app may also need microphone permission. The exact shortcut can be changed in System Settings > Keyboard > Dictation, so an old tutorial may show a trigger that no longer matches your Mac.

The Three Voice Modes You Need to Know

Apple's current Voice Control documentation distinguishes Dictation mode, Spelling mode, and Command mode. These modes solve different input problems, so confusing them creates many of the “why did my Mac type that?” moments.

Dictation mode accepts ordinary speech and places recognized words at the cursor. In Notes, you might say, “The design review starts tomorrow period,” and receive a sentence with punctuation. Apple states that Dictation mode is enabled by default.

Spelling mode accepts individual characters rather than normal words. It's intended for difficult strings such as passwords, web addresses, and character sequences. Apple gives the phonetic example “Alfa Bravo Charlie,” which produces “abc.” This mode is safer for exact text because each spoken code word represents a character.

Command mode interprets speech as an action. Instead of inserting the words “select previous sentence,” Voice Control uses that phrase to select text, provided the command is supported in the current context. Commands can access, select, delete, format, or activate interface controls.

The boundary is intentional. If you say a command while the Mac is listening for prose, it may insert literal words. If you dictate a sentence while Command mode is active, the system may search for matching actions instead of writing the sentence.

Switching modes is itself part of the voice-control model. That's why reliable workflows group commands by intent, then choose the mode that can interpret the utterance correctly.

Built-In Formatting and Navigation Commands

The built-in command vocabulary is most useful when grouped by the job you're trying to complete. Apple documents punctuation, capitalization, line control, spacing, and navigation commands as part of Mac dictation rather than treating dictation as transcription alone.

Punctuation and capitalization

Say punctuation as part of the sentence:

• “Period” ends a sentence. Example: “The deployment is complete period”
• “Comma” inserts a comma. Example: “After testing comma we'll publish the notes”
• “Question mark” closes a question. Example: “Can you approve the change question mark”
• “Exclamation point” adds emphasis. Example: “The backup finished exclamation point”
• “Colon” introduces a list or explanation. Example: “Bring three items colon notebook comma charger comma adapter”
• “Semicolon” separates closely related clauses. Example: “The test passed semicolon the warning remains”
• “Dash” inserts a dash where supported. Example: “The result dash not the cause”
• “Ellipsis” inserts an ellipsis. Example: “I thought we had more time ellipsis”

Capitalization commands help when ordinary recognition doesn't preserve the form you need:

• “Cap” capitalizes the next word. Example: “cap monday”
• “All caps” produces uppercase text. Example: “all caps urgent”
• “Caps on” keeps subsequent dictated text uppercase until you say “caps off”
• “No caps” prevents capitalization for the next entry where supported

Spacing and structure

Use “new line” when you want the cursor to move to the next line. Use “new paragraph” when the text should begin a separate paragraph. The difference matters in documents where line breaks and paragraph breaks carry different formatting.

Apple also lists “no space on” and “no space off.” These commands help with strings that must remain joined. For example, you can say “no space on project code no space off” when a normal phrase needs to be entered without an intervening space.

Navigation and correction

For movement and cleanup, Voice Control commands are often more appropriate than basic Dictation mode. Try “move to end of sentence,” “select previous word,” or “undo that.” If the command is inserted as text, switch to Command mode and repeat it.

For a broader punctuation reference with practical examples, see this list of punctuation commands for dictation.

Treat the stock command list as a reliable foundation, not a complete voice-operated editing environment. It handles common structure well, but exact symbols, application actions, and advanced rewriting often require Spelling mode, Voice Control, keyboard input, or another layer.

Spelling Mode and Phonetic Alpha Bravo Charlie Entry

Natural dictation is designed to recognize words. That's precisely why it struggles with passwords, URLs, access tokens, filenames, and code fragments. Spelling mode changes the task from “understand my sentence” to “enter these characters.”

Start spelling with the voice command available in your Voice Control setup, or use the documented keyboard activation pattern when supported. Apple's current guidance explains that Spelling mode is intended for difficult strings and gives “Alfa Bravo Charlie” as a phonetic example that produces “abc.”

A practical session looks like this:

1. Activate dictation and wait for the ready signal.
2. Enter Spelling mode.
3. Speak phonetic entries for letters.
4. Say character names for symbols when supported.
5. Return to natural dictation with “resume dictation.”

The phonetic alphabet is useful because ordinary letter names can blend together. Examples include Alfa for A, Bravo for B, Charlie for C, Delta for D, Echo for E, Foxtrot for F, Golf for G, Hotel for H, India for I, Juliet for J, Kilo for K, Lima for L, Mike for M, November for N, Oscar for O, Papa for P, Quebec for Q, Romeo for R, Sierra for S, Tango for T, Uniform for U, Victor for V, Whiskey for W, X-ray for X, Yankee for Y, and Zulu for Z.

“Spell that” asks the system to re-enter the previous token character by character. “Letter” can introduce a single character, while capitalization modifiers can help preserve uppercase entries. For a technical string, don't mix a long natural-language sentence with exact symbols. Speak the exact segment in Spelling mode, then say “resume dictation” before continuing prose.

Stock Dictation Compared to AIDictation

Apple's built-in dictation and a third-party tool such as AIDictation occupy different layers. Stock macOS dictation gives you system-level input, Apple's command vocabulary, Voice Control modes, custom vocabulary support, and workflows that can operate across text fields. A third-party layer can focus on transforming rough speech into a finished document.

Consider three workflow moments:

AIDictation's stated product model includes Auto Mode, which can choose between on-device recognition and a cloud service, and Local Mode, which runs on Apple Silicon without an internet connection. Its cloud-connected workflow adds cleanup, contextual formatting, filler-word removal, and handling for self-corrections. Those capabilities address a different stage of work from Apple's command interpreter.

The integration is additive rather than a replacement model. Built-in phrases such as “new line” and “delete that” remain part of the macOS workflow, while a third-party layer can interpret higher-level instructions such as inserting a template, reshaping a note, or formatting a code comment.

Use stock commands when you need direct control over the cursor and document. Consider an additional layer when the main problem is not entering words, but turning rough speech into a polished email, clinical note, specification, or developer explanation.

Custom Dictionary and App-Specific Context Rules

Apple's custom vocabulary is useful for names, product terms, and specialized words that the speech engine repeatedly misrecognizes. Open System Settings > Keyboard > Dictation, then look for the custom vocabulary controls available on your macOS version. Add the preferred spelling, test it in a simple text field, and remove entries that create new errors.

A plain vocabulary entry treats a term as a general replacement. That's different from an app-scoped rule. The same spoken shortcut might reasonably expand differently in an electronic health record, a code editor, Mail, or a chat window.

Separate vocabulary from context

A clinical workflow could use a trigger such as “clinical note” to insert a structured prompt or template in the active clinical application. A developer might use “code comment” to produce a concise technical explanation in an editor. These rules should be scoped carefully, because expanding “pt” to “patient” everywhere could make an ordinary message awkward, while expanding “cb” to “callback” may be useful only in source code.

A context layer such as AIDictation can add app-scoped rules, regex-style matching, and prompt-driven formatting. The practical setup is:

1. Add universal names and terms to Apple's custom vocabulary.
2. Create a distinct trigger phrase for each specialized workflow.
3. Scope the rule to the relevant application.
4. Dictate a short test sentence.
5. Toggle the rule off if it produces an unexpected rewrite.

The Mac custom dictionary guide is useful when you need to distinguish plain vocabulary from context-aware replacements.

Review custom rules periodically. A replacement that helps in one application can become disruptive when the same trigger appears in another, and accumulated vocabulary can drift as your work changes.

Keyboard Shortcuts and the Fn Twice Trigger

The most dependable activation method is the one your hands remember. Open System Settings > Keyboard > Dictation and choose the available shortcut that matches your keyboard. Apple also supports starting dictation with the Microphone key, a keyboard shortcut, or Edit > Start Dictation.

Some Mac users still expect the older Fn twice gesture. Apple's current documentation acknowledges the historical use of pressing the Fn key twice, but shortcut availability depends on macOS settings, keyboard hardware, and whether another accessibility feature has claimed the gesture.

If the default trigger feels awkward, choose a custom shortcut in the Dictation settings. An external keyboard may lack a Globe or Microphone key, so a configured key combination can make activation more consistent across machines. The cursor still needs to sit in an editable text field.

Use Edit > Start Dictation when a menu command is easier than reaching for a key. This fallback is especially useful with a trackpad-only setup or when you're testing whether the problem is the shortcut rather than the speech engine.

Voice Control can change shortcut behavior because it uses its own listening and command model. If Fn twice doesn't respond, check whether Voice Control is enabled, then verify the active Dictation shortcut instead of assuming the gesture is still configured.

For a wider hands-free control workflow, see this guide to Voice Command on Mac.

App-Specific Commands Across Mail, Notes, Pages, and Editors

System-wide commands usually act on the active text field. App-specific commands depend on the frontmost application and the controls it exposes. That's why the same phrase can insert text in one place and perform an action in another.

“New line” generally changes document structure, but the exact visual result depends on the editor. “Send message” is a clear example of context. In Messages, it can submit the conversation. In Mail, it can send the current draft. In a plain Notes field, it may be treated as text unless Voice Control recognizes an available action.

Selection, deletion, and undo are broadly useful because they operate on text or the latest edit. Sending, inserting checklist items, changing page position, and jumping to a source line are more dependent on app-specific commands.

When you're unsure what an app supports, use Voice Control's Show Commands overlay. It exposes available phrases for the current environment instead of forcing you to guess whether a command is global or local.

Example Utterances for Clinical, Developer, and Multilingual Workflows

A command recipe works best when it separates content from control. Speak the prose naturally, announce structure explicitly, and reserve exact characters for Spelling mode.

Clinical documentation

For a SOAP-style note, dictate each section as a separate block:

“Subjective colon patient reports improved sleep period new paragraph Objective colon blood pressure stable period new paragraph Assessment colon symptoms improving period new paragraph Plan colon continue observation period”

Use capitalization only when a name or identifier requires it. For a workflow that produces formal clinical documentation, a medical scribe may also help separate raw conversation capture from the final record, subject to your organization's privacy and compliance requirements.

Developer notes and comments

Use natural language for the explanation, then switch to exact entry for identifiers:

“Open brace new line This function validates the request period new line go to end of line semicolon”

For a variable with unusual capitalization, enter the name in Spelling mode rather than asking ordinary dictation to infer it. Commands such as “open brace” and “close brace” may depend on the active Voice Control command set, so verify them in the editor before relying on them.

Multilingual writing

Change the active dictation language through your macOS language settings or the shortcut configured for language switching. Then speak the content in the selected language, while keeping structural commands in the form recognized by the active Voice Control configuration.

A starter sequence might be:

“Start dictation [switch language] Bonjour, je vous envoie le document period new paragraph [switch language] The revised summary is attached period”

If autocorrection repeatedly changes an accented name or technical term, use a literal or spelling entry where available, then return to normal dictation. Test multilingual switching in Notes before using it inside a form that might submit text when you press Return.

Why Enhanced Dictation and Custom Dictation Commands No Longer Apply

Many older Mac guides send readers to Enhanced Dictation and a separate Custom Dictation Commands panel. That advice describes an earlier organization of Apple's speech features, not the modern path most users see.

Apple's current architecture separates Dictation, Voice Control, and vocal shortcuts. Voice Control provides the structured interaction model with Dictation, Spelling, and Command modes. Its documentation also describes speech commands for navigation, editing, and custom actions, which is why current troubleshooting should begin in Accessibility > Voice Control when the problem involves commands rather than simple text entry.

The old “insert a phrase” workflow may still appear in archived instructions, but it shouldn't be treated as the primary configuration route. Look for current controls related to Voice Control commands, vocabulary, languages, and recognition behavior instead of searching for a legacy Enhanced Dictation checkbox.

Rule of thumb: Use Keyboard settings to activate Dictation, and use Accessibility > Voice Control to manage spoken actions and command behavior.

Third-party tools can fill a different customization gap by applying context rules, cleanup, or specialized templates after speech is captured. That doesn't make Apple's built-in commands obsolete. It clarifies which layer owns activation, transcription, navigation, and rewriting.

Troubleshooting the Most Common Dictation Failures

Most dictation failures have a single likely cause. Work from the symptom rather than changing several settings at once.

The ready-state signal is a useful diagnostic. If the indicator appears but the Mac inserts the wrong content, activation worked and the issue is likely mode, language, vocabulary, or command scope. If the indicator never appears, focus on permissions, the shortcut, the active field, and competing voice features.

When a phrase is inserted, don't repeat it louder. Change the interpretation state first. Natural prose belongs in Dictation mode, exact characters belong in Spelling mode, and navigation or editing belongs in Command mode.

Quick-Reference Card for Mac Dictation Commands

Keep this compact card beside your Mac until the commands become automatic.

Activation

• Microphone key: Start dictation when the cursor is in an editable field.
• Configured keyboard shortcut: Use the shortcut selected under Keyboard > Dictation.
• Fn twice: Available on supported configurations and documented as a historical Mac trigger.
• Edit > Start Dictation: Menu fallback when the keyboard trigger is inconvenient.

Structure and punctuation

• “Period”: Insert a full stop.
• “Comma”: Insert a comma.
• “Question mark”: Close a question.
• “Exclamation point”: Add an exclamation mark.
• “New line”: Move to the next line.
• “New paragraph”: Start a separate paragraph.
• “No space on” / “no space off”: Control spaces around dictated text.

Capitalization and exact entry

• “Cap”: Capitalize the next word.
• “All caps”: Enter the next text in uppercase.
• “Caps on” / “caps off”: Toggle uppercase entry.
• “Start spelling”: Enter character-by-character text.
• “Alfa Bravo Charlie”: Produce the letters abc.
• “Resume dictation”: Return to natural-language transcription.

Editing and app context

• “Undo that”: Reverse the latest supported edit.
• “Select previous word”: Select text before the cursor.
• “Move to end of sentence”: Move within the current text.
• “Send message”: May send a draft in Mail or a conversation in Messages.
• “Next page”: May move through Pages.
• “Go to line forty-two”: May move within a developer editor.

For more advanced formatting, use context rules that distinguish Mail, clinical software, chat, and code editors. Test every rule in a safe draft before using it in a live form.

---

AIDictation adds AI cleanup, context-aware formatting, custom vocabulary, and local or cloud recognition options to the foundation provided by macOS dictation commands. If you want to turn rough spoken input into structured emails, notes, clinical drafts, or technical documentation, visit AIDictation and test how its workflow fits your Mac.