Privacy Policy
Last updated: August 12, 2026
Dictaste (“we”, “us”) provides desktop dictation apps for macOS and Windows and related web services. This policy explains what we collect, what stays on your device, and how free, developer, and paid cloud features work.
Speech processing
Mac: core speech-to-text uses Apple's on-device Speech frameworks. Microphone audio for transcription is processed locally. We do not operate a server that receives your raw audio for Mac dictation. Optional Apple Intelligence polish (when available) follows Apple's on-device terms.
Windows: default STT uses the browser Web Speech path in the app (online, provider-dependent). You may switch to OpenAI Whisper or a local whisper.cpp binary with your keys/models. We do not receive raw Windows mic audio unless you explicitly use a managed cloud path we host (we currently do not for STT).
Highlight-to-speak
When you select text, Dictaste may read that selection aloud. System voices process speech on-device (Mac AVSpeech · Windows SAPI). If you enable premium voices (managed Pro TTS or your own OpenAI key), only the selected text is sent to the TTS provider—never microphone audio. Selection capture stays on your device (Mac Accessibility / Windows hotkey + clipboard restore).
Your own API key (free for developers)
If you add your own OpenAI or other provider key in the app (free developer path), polish requests send text only (not audio) from your device to that provider under your account. Your API key is stored only on your device. We never receive or store it. New installs and “Reset all settings” clear keys and return to defaults.
Managed AI polish (Free, Pro, Pro Plus)
When you use managed polish (Free daily allowance or a paid plan), the Mac app sends the text transcript (not microphone audio) to our API with your license key. We forward that text to model providers we use to generate cleaned text—primarily NVIDIA NIM and, if configured as fallback, OpenAI. We meter word usage against your plan and return the result. We do not sell your dictation content. Provider processing is subject to their respective privacy terms.
Account & billing data
When you create an account we process email and profile data via our auth provider (Clerk). Payments are processed by Stripe; we store subscription status, Stripe customer id, plan, usage counters, and license keys needed to unlock features. We do not store full payment card numbers.
GitHub star (Developer plan)
To unlock the free Developer plan we may verify that a GitHub username you provide has starred our public repository. We store your GitHub username and a starred flag on your account for entitlement checks. We do not store your GitHub password or tokens.
Affiliate program
If you join the affiliate program we store your referral code, referred user attributions, and commission ledger entries needed to calculate payouts. Payouts may require additional tax or payment details you choose to provide.
Website
Our marketing site may use standard server logs (IP address, user agent, path, timestamps) for security, abuse prevention, and debugging. Cookies may be set by Clerk for authentication. We do not sell personal data.
Data retention
Account and billing records are kept while your account is active and as required for tax, legal, or dispute purposes. Usage counters reset by plan period (daily or monthly). You may request deletion of your account by emailing support.
Children
Dictaste is not directed to children under 13. We do not knowingly collect personal information from children under 13.
Your rights
Depending on your location you may have rights to access, correct, export, or delete personal data we hold. Contact us to exercise these rights. If you use Clerk or Stripe, some data is controlled by those processors under their policies.
Contact
Privacy questions: GitHub Issues.