Dictation Not Working on Mac? Check These Five Things

If dictation starts but produces no text, first check the microphone and the field you are trying to type into. If it never starts, check the trigger. Reinstalling a dictation app before separating those two cases makes it harder to tell what fixed the problem.

1. Identify which dictation tool is failing

macOS Dictation, a website’s microphone button, and Notchy’s Dictate Text command are separate tools. A change in Keyboard settings affects Apple’s built-in Dictation; a browser permission affects that website. Begin with the tool you actually invoked, rather than changing every microphone setting at once.

For built-in Dictation, open System Settings → Keyboard and find Dictation. Check that it is enabled and note the configured shortcut. Apple documents the activation options in its Mac Dictation guide. Setting labels can vary with your macOS version.

2. Check the microphone and recognition language

A connected headset is not necessarily the input you intended to use. In System Settings → Sound → Input, select the microphone you want and speak a short sentence. Check whether its input level responds. Try the built-in microphone as a comparison if a Bluetooth headset gives no useful signal.

For Apple Dictation, also inspect the microphone source and language in the Dictation settings. Select a language you are actually speaking and test one short sentence before a long paragraph. Apple’s guide explains these controls; a wrong language and a silent microphone require different fixes.

3. Separate a shortcut problem from a typing problem

Open a new note, click in the text area, and start dictation from its menu or command rather than your usual hotkey. For Apple Dictation, use Edit → Start Dictation where available. For Notchy, open the command palette and run Dictate Text.

If the command works but the hotkey does not, inspect the shortcut assignment and conflicts. If text works in Notes but not in the original app, inspect the original field: it may be read-only, lack keyboard focus, or handle pasted content differently. Test a normal sentence, not a password or an authentication code.

4. Check Notchy’s permissions and local model

Notchy uses Apple’s on-device speech recognition and pastes the finished transcript into the app that had focus. Check its microphone and speech-recognition access in System Settings → Privacy & Security, and follow any permission prompt for inserting text. Our permission guide explains why different tools ask for different access.

Notchy follows the system language and requires an available offline speech model. If it reports that a local model is unavailable, a working microphone alone will not resolve that message. See the dictation feature page for the supported behavior; Notchy does not silently substitute cloud recognition.

5. Test a short, repeatable workflow

1. Place the cursor in a new note and run Dictate Text.
2. Say one sentence and watch the notch’s recording activity and word count.
3. Run the command again to stop, then check the inserted transcript.
4. Repeat in the destination app before trying a longer passage.

Notchy also ends a session after a pause in speech, so prepare what you want to say before starting. Use a short first test to distinguish recognition from insertion. If neither app recognizes speech with the built-in microphone, return to the system input and recognition settings before changing Notchy again.

For text that already exists in an image, copy it with OCR instead of reading it aloud. Dictation is most useful for a new thought; OCR is better suited to preserving an existing message.