VoiceInk privacy: what stays on your Mac, what is sent, and the three settings to check

By Ryleigh Newman · Published October 10, 2026

VoiceInk can run entirely on your Mac, and it can also send your words to a cloud provider. Which one happens depends on three separate settings: the transcription model, AI enhancement, and Dictionary Auto Learn. Its docs say to check all three, because a local transcription model does not stop the other two from sending text out.

I built YapToText, a different dictation app for the Mac, so read what I say about VoiceInk with that in mind. Every fact about VoiceInk here is from its documentation, read on October 9, 2026. The pages are Privacy and Data, Context Awareness and Installation.

This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.

This guide in other languages: العربية, Čeština, Dansk, Deutsch, Ελληνικά, Español, Suomi, Français, עברית, हिन्दी, Magyar, Bahasa Indonesia, Italiano, 日本語, 한국어, Norsk, Nederlands, Polski, Português, Română, Русский, Slovenčina, Svenska, ไทย, Türkçe, Українська, Tiếng Việt, 简体中文, 繁體中文

What each feature uses

From VoiceInk's docs page Privacy and Data, read October 9, 2026.

The fully on-device setup

1. Select a local transcription model in the Mode you use.
2. Turn AI enhancement off, or use a local option. VoiceInk Refine runs locally on an Apple silicon Mac with at least 16 GB of memory.
3. Turn Dictionary Auto Learn off, or point it at a local review model.
4. Repeat for each Mode.

Local transcription does not make cloud enhancement local. That sentence is close to the docs' own wording, and it is the thing to remember. The VoiceInk Modes page covers where each setting lives.

Why it asks for Screen Recording

From VoiceInk's docs page Installation.

Screen context is optional. When it is on, VoiceInk captures the active window, reads the text from it on your Mac with Apple's Vision framework, and sends that extracted text to the AI provider as part of the prompt. The docs say the screenshot image itself is not sent. If you never grant Screen Recording, screen context is skipped and dictation still works.

What it reads when recording starts

At the start of a live recording VoiceInk prepares a snapshot in memory: the clipboard, the selected text, and, when permission allows, text from the active window. A Mode's switches for Selected Text, Clipboard and Screen decide which of those are included in the enhancement request. The docs are precise about this: turning a source off stops it from being sent, and does not stop every local read. Plain dictation with no enhancement sends none of it to a provider.

What is kept on your Mac

• History. Transcripts, enhanced results, saved audio and AI request details are stored locally. The request details can include the context that was sent.
• Deleting. History settings control how long transcripts and audio are kept. The docs note that deletion is not a guarantee of forensic erasure from the database, backups or the disk.
• API keys. Stored in the macOS Keychain, and left out of settings backups.
• Cloud services. Each one applies its own data and retention policy.

YapToText next to it

Both apps can keep everything on the Mac. The difference is how much you have to check.

VoiceInk from its docs, read October 9, 2026. YapToText from its own source and settings.

VoiceInk gives you more models and more control. YapToText has no cloud option to misconfigure. The private dictation page shows how to check any app with the network off, and the VoiceInk alternative page compares the two overall.

This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.

The short version

• Local. A local transcription model keeps the audio on your Mac.
• Check. AI enhancement and Auto Learn, separately.
• Screen Recording. Optional. It sends extracted text, not the screenshot.
• History. On your Mac.
• Open source. You can read what it does.

If VoiceInk changes how it handles data, open an issue on GitHub and I will update this page. Every one gets read.

This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.

About the author. Ryleigh Newman is a developer whose hands make typing difficult and who dictates every day with YapToText, the app this site describes. Facts about other apps come from their own sites, with the date they were read. More about the project.

Related

• VoiceInk guides: every page on this site about VoiceInk, by topic
• VoiceInk Dictionary Auto Learn: how it learns from your corrections, and what it shares
• VoiceInk cloud providers: which ones transcribe, how to add an API key, and who pays
• VoiceInk Context Awareness: what it reads from your Mac, and where that text goes
• VoiceInk History: finding old transcripts, exporting them, and deleting them on a schedule
• The VoiceInk Dictionary: vocabulary, word replacements, Auto Learn and filler words
• VoiceInk Refine and AI Enhancement: clean up dictation locally, and when you need a provider
• VoiceInk Modes and triggers: how it switches behavior by app, website and spoken word
• Which VoiceInk model to choose, from VoiceInk's own guidance
• A free VoiceInk alternative, open source with nothing to compile

Sources

• https://tryvoiceink.com/docs/privacy-and-data (checked 2026-10-09)
• https://tryvoiceink.com/docs/context-awareness (checked 2026-10-09)
• https://tryvoiceink.com/docs/installation (checked 2026-10-09)
• https://tryvoiceink.com/docs/voiceink-refine (checked 2026-10-09)
• https://github.com/ryleighnewman/YapToText (checked 2026-10-10)
• https://apps.apple.com/us/app/yaptotext/id6786382289 (checked 2026-10-10)

Questions