Your audio never leaves your Mac — in any plan. Free is fully local. Pro AI features use zero-retention cloud infrastructure because it delivers more reliable results than running on-device. Either way, nothing you say is ever stored or used for training.
Speech recognition runs entirely on your Mac using local models. No account required, no network traffic, no data of any kind leaves the device. Your dictionary and settings are stored locally in your home directory.
STT is still local. AI post-processing — formatting, style, memo insights, Dispatch — sends your transcript text (never audio) to Groq via our server proxy. Groq operates with zero data retention: your text is not logged, stored, or used to train any model.
Raw audio is captured in memory, transcribed locally, and discarded immediately. No audio file is ever transmitted — on any plan.
Analytics events (word counts, feature usage) never include the content of what you dictate. Transcription text is explicitly scrubbed from crash reports before they leave the device.
We sell software, not data. Nothing you say or write is used to train AI models, shared with advertisers, or sold to third parties.
All STT models run on your Mac's Apple Silicon chip. Audio is processed in real-time and discarded after transcription. This applies to every plan — Free and Pro alike. The models are bundled with the app and optimized for M1 through M5.
When AI post-processing is active (Pro), your transcript text is sent to our server, which forwards it to Groq for cleanup, style matching, and formatting. Groq processes requests under a zero data retention agreement — meaning your text is not logged, not stored, and not used to train any model. We use cloud processing here because it produces more consistent, higher-quality results than running equivalent models locally.
What's sent: transcript text, window title (for tone context), your dictionary terms, and a short style hint derived from your correction history. What's never sent: audio, screenshots, or any content you haven't dictated.
Your personal dictionary lives locally on your Mac. Correction examples — short snippets of text you've edited after pasting — are stored in your Resonant account on our server so they persist across reinstalls and are available for style matching. These examples are capped at 120 characters each and are only used to improve your own output, never anyone else's.
Relay connects your phone to your Mac over local WiFi using encrypted connections. All traffic stays within your local network. No audio leaves the network, and no relay server or cloud infrastructure is involved.
Last updated: March 9, 2026
Name: Thomas Schlossmacher
Business form: Individual entrepreneur (Einzelunternehmer)
Country: Austria
Contact: team@onresonant.com
Resonant is designed to collect as little data as possible. We never collect audio recordings. Transcript content is never included in analytics or crash reports.
Depending on which plan you use, the following data may be collected:
Transcript text processed by Groq is not retained after the response is returned (zero data retention). Personalization examples are retained as long as your account is active and deleted upon account deletion. Account information is retained for the duration of your subscription and deleted upon request.
Under the GDPR and Austrian data protection law (DSG), you have the right to:
To exercise these rights, contact us at team@onresonant.com.
We use the following sub-processors:
If you believe your data protection rights have been violated, you have the right to lodge a complaint with the Austrian Data Protection Authority (Österreichische Datenschutzbehörde) at dsb.gv.at, or with the supervisory authority in your EU/EEA country of residence.
We may update this Privacy Policy from time to time. Material changes will be communicated through the application or this website. Your continued use of the Service after changes constitutes acceptance of the updated policy.
Privacy isn't a feature we added. It's how we built it.
Download Resonant