Troubleshooting

Recording will not start

1. Check Microphone access under Settings → Permissions.
2. Select a connected input under Settings → Voice → Recording → Microphone.
3. Try Start dictation in the menu-bar panel. If it works, check the shortcut and Input Monitoring for a modifier-only shortcut.
4. If the app reports a trial or licence issue, open Settings → Licence.

If MicMyDay is missing from a macOS permission list, use the permission drag helper. If a shortcut is already in use, choose a different combination.

AirPods take a moment to start

A brief delay with AirPods or another Bluetooth headset is expected: the audio hardware wakes up and macOS switches the connection into microphone mode. MicMyDay cannot bypass that hardware step. Wait for the spinning ring and Starting… to finish before speaking, even if you have already heard the opening tone. See the startup indicator and why it appears.

If the delay keeps going, or the input level stays flat when you speak after the spinner disappears, stop the attempt, check that the headset is connected, and select it again under Settings → Voice → Recording → Microphone. Try reconnecting the headset or choosing a wired or built-in microphone, then record a short sentence. For persistent silence, use the audio checks below.

The recording is silent or has no speech

Check that the selected microphone is connected and unmuted. Speak while watching the input level, and verify the input under System Settings → Sound → Input.

If the app reports no usable microphone or no audio frames, try another input. If it recorded audio but could not recognize speech, speak closer to the microphone and try a short, clear sentence. See Microphone & audio.

Text is copied but not pasted

Press ⌘V to use the result now. For the latest result, you can also use Last transcription → Copy transcription in the panel.

Check Settings → General → Paste automatically and Accessibility permission. Restart MicMyDay after allowing Accessibility. Click the destination text field before starting the next recording.

If a partial result was typed, follow the panel’s instruction for either pasting the remainder or replacing it with the full text. See live-text recovery.

A local model is not ready

Select an installed model, or use Get to download the model you want. Wait for downloading and preparation to finish. Speech models are under Engine; rewrite models are under Rewrite → On this Mac → Model.

For repeated loading failures, delete that model using its trash control and download it again. Use a working internet connection for the download. For a large model or long rewrite, try a smaller model or shorter input.

Words or language are wrong

• Check the language coverage of your speech model.
• Try an explicit Spoken language with Whisper or Nemotron if detection is wrong. Parakeet always detects the language itself. With Apple Speech, Automatic uses your Mac’s current language and region.
• Save a correction for a recurring misspelling.
• If words disappear, review the filler list. An entry may be meaningful in your language.
• If the pre-rewrite text is right but the result is wrong, adjust the rewrite profile or turn rewriting off to compare.

Rewriting is unavailable or gives an unexpected result

Check that rewriting is enabled, its chosen model is ready, and the intended profile is selected. Your selected profile stays active until you choose another one.

For Apple built-in, check Apple Intelligence, macOS version, eligibility, model readiness, and language support. Downloaded local rewrite models have different requirements. See Rewriting on your Mac.

If a rewrite fails, the text before that rewrite is kept. If the result has the wrong format, shorten or clarify your prompt using the prompt-writing suggestions.

A provider or custom server fails

Check the key, selected model, account access or credit, and network connection. Refresh the model list when available. A model offered by a provider may not yet be supported by the installed app version.

For Custom, check the base URL, authentication, and the endpoint required for audio transcription or chat rewriting. A server on another local device may need Local Network permission. See Custom endpoints.

For a request that is too large, record a shorter segment and try again. If the issue continues, try another configured model to narrow down the cause.

Live text does not appear

Check whether the provider and model support streaming. For overlay text, the overlay must be enabled and set to Compact or Wide. Direct typing may use the overlay instead when cleanup needs to run first. See Live text.

Trial, licence, or update issues

An expired trial requires activation. Licence check needed requires a successful online validation; reconnect and reopen the app. For an unrecognised key or device-limit error, follow Licence & trial.

For updates, check your internet connection and use Settings → General → Updates → Check for Updates. If the updater reports an error, include it when contacting support.

Send a useful support request

Contact support with:

• Your macOS version and MicMyDay version/build, shown under General → Updates → Currently.
• Your transcription engine and model, plus rewrite provider/model if relevant.
• The exact error message and what you were trying to do.
• Whether the problem repeats, and the checks you have already tried.

Include a short example if it helps explain the issue, with personal details removed. Keep API keys and licence keys private.