Is voice dictation private? Does your audio get uploaded?

On-device vs cloud processing

• On-device recognition: audio is turned into text on your device and doesn’t need to leave it.
• Cloud processing: audio is uploaded to the provider. Then it comes down to what they promise: encrypted transfer, deletion after processing, whether it’s used for training, and which certifications they hold.
• AI cleanup is a separate step. Even with on-device recognition, text sent to a cloud model for cleanup leaves your computer. Know whether it’s audio or text that leaves, and who receives it.

What each product states

Competitor information checked on 2026-11-15; their official sites are authoritative (sources at the end). Qugora prices are planned.

Check whether Mac Dictation runs on your device

1. Choose Apple menu > System Settings, then click Keyboard in the sidebar.
2. With Dictation on, read the text below Dictation. It tells you whether voice input and transcripts for general text dictation (composing messages and notes, not dictating in a search box) are processed on your device and not sent to Siri servers.
3. If you don’t want Apple to store audio of your dictation, go to Privacy & Security > Analytics & Improvements and turn off Improve Siri & Dictation.

How Qugora handles your data

• Speech recognition runs on device; audio and text stay on your computer by default, and basic voice input works offline.
• AI cleanup and candidates can use Qugora Cloud (online; only token counts are recorded for billing, request content isn’t stored), your own API key (text goes straight from your device to the provider you chose, not through us), or a local model.
• Telemetry is counts and timings only, never text or audio, and can be turned off.
• Memory items need your approval one by one; only approved items sync, and only if you turn on sync (Pro).
• Your audio, text and memory items are not used to train models.

See the privacy policy (draft). The Qugora macOS preview (Apple silicon) is free to download on the Download page; the Windows build and Pro subscriptions aren’t open yet.

A privacy checklist for any voice input tool

1. Is speech recognized on device or in the cloud? Look for a clear statement in the official privacy docs.
2. If it’s in the cloud, how long is audio kept, and can you turn storage off?
3. Is your data used for training, and can you opt out?
4. Whose model does the AI cleanup, and can you use your own API key or a local model instead?
5. Does telemetry include text or audio, and can you turn it off?