Turn a thought into text.
Keep speech on your Mac.

How to dictate into a Mac app with Padhow

1. Open Padhow Settings → Dictation. Download a speech model, select it and save your settings.
2. Grant Microphone access for recording and Accessibility access for shortcuts and insertion. Try the text area in Dictation settings first.
3. Focus an editable text field. Hold Right Option (⌥), wait for the listening indicator, then speak. Release to finish.
4. Padhow transcribes locally and saves the text to Voice clipboard before attempting insertion. Review the inserted text before sending it.

For example, dictate a reply to a colleague or a first draft of a note. If you also want to shorten, translate or rewrite it, use a separate AI writing action after dictation.

Default dictation shortcuts

• Hold Right Option: hold to talk, release to finish.
• Control+Option+Space: start or stop hands-free recording.
• Escape: cancel the active dictation.
• Control+Option+H: open Voice clipboard.
• Control+Option+V: paste the last transcript into the current field.
• Control+Option+C: copy the last transcript.

Change shortcuts in Dictation settings. Hold-to-talk also supports Fn / Globe or a custom combination. Left Option does not activate the default Right Option gesture.

Mac dictation requirements and limits

Padhow’s speech helper requires macOS 14 or later. The current website download is for Apple Silicon. Speech-model options include Whisper and Parakeet; Parakeet requires Apple Silicon. Internet access is needed to download a model, then transcription runs locally.

Speech typing currently supports English. Recording stops after five minutes. Secure keyboard entry blocks global dictation, and password fields are not insertion targets. App support depends on macOS Accessibility exposing an editable field.

Apple Intelligence is not required for speech typing or for using Padhow with another supported AI model. Optional Apple Intelligence writing help requires macOS 26 or later, compatible hardware, and its model enabled and downloaded. See Padhow compatibility.

Where do the recording and transcript go?

Padhow’s speech recognition runs on-device. Live dictation does not call a chat model or automatically fall back to cloud transcription. Temporary live audio is discarded after processing or cancellation. Voice clipboard stores transcripts locally, separately from ordinary clipboard history.

Completed dictations remain in Voice clipboard until you delete them; there is no automatic expiry. You can inspect, copy or delete entries there. Saved dictionary corrections and voice snippets can change the transcript without sending it to an AI service.

If you later submit text to a cloud writing provider, that provider receives the text and context you choose. Its pricing and data policies apply. Read about local and cloud AI options.

Why was my dictation saved but not inserted?

Padhow only inserts when the original editable field, selection and text are unchanged. If focus moved, permissions are missing or the field is unsupported, recover the transcript from Voice clipboard and paste it yourself. You can also dictate with no text field focused to save directly to Voice clipboard.

Can I personalize names and repeated phrases?

Add preferred words and explicit corrections in Dictation settings. Create a voice snippet for a repeated phrase and say its trigger by itself to insert the saved text. This is dictionary and snippet customization, not model training.

Does local dictation have a subscription?

Padhow’s app price is €12 once for one person on all their Macs, including future app updates. There is no Padhow subscription or per-minute transcription fee. Optional cloud AI usage is billed separately. The Mac download is available; public checkout is being prepared.