Privacy Policy
Austin's Transcriptor is built around a simple idea: your audio is yours, and it should never leave your Mac. The app does its work entirely on your device.
What the app does on your Mac
You drag an audio file (or capture system audio) into the app. The app converts it to PCM, hands it to a local Whisper model, and produces a transcript. The model runs inside the app's sandbox on your machine. Your audio is never uploaded.
If you use AI Transcript Tools, the transcript is fed to a local Gemma 4 model running inside the same app sandbox. The transcript, your AI prompt, and the AI response all stay on your device. Nothing is uploaded.
Transcripts are stored as JSON files on your Mac at ~/Library/Application Support/AustinsTranscriptor/transcripts/. Recordings the app captures (Microphone, System Audio, Mic + System) are saved as .wav files at ~/Library/Application Support/AustinsTranscriptor/audio/ so you do not lose them if anything goes wrong. Imported files (drag-and-drop, YouTube) stay where you put them and are referenced via Apple's security-scoped bookmarks. You can delete an individual recording from its transcript, or clear every saved recording from Settings → Storage.
What our servers see
None of your audio. None of your transcripts. We don't run servers that touch your content.
The app downloads model files from Hugging Face on first use. (1) On first launch the app pulls OpenAI's Whisper Large v3 Turbo weights (1.55 GB, from ggerganov/whisper.cpp). (2) The first time you enable AI Transcript Tools, the app pulls Google's Gemma 4 language model (about 5.2 GB) for the on-device summary and chat. Hugging Face receives standard HTTP metadata for those downloads (your IP, user agent). After download, both models live on your Mac and the app no longer talks to Hugging Face for them. Neither model phones home or transmits your audio, transcripts, or any other content.
If you import a video by pasting a YouTube link, the app sends that URL to a third-party YouTube extraction service (remote-production.youtubekit.dev) to resolve the audio stream. Only the URL is sent; the resulting audio downloads from YouTube directly to your Mac and is transcribed locally. If you'd rather not contact a third-party service, save the audio in your browser and drop the file in instead.
When you make the $39.99 in-app purchase or use Restore Purchases, the app validates the receipt through Apple's StoreKit. This may involve a network call to Apple's servers as part of standard Apple in-app purchase plumbing; we never see the contents of that exchange.
This marketing website (austinstranscriptor.com) is hosted on Netlify and protected by Cloudflare. Standard request metadata is visible to those providers. We do not run analytics on this site.
Advertising
The app shows no advertising. No banners, no inline promos, no third-party ad SDKs. The free tier is capped at 5 transcriptions per local calendar day (file imports, YouTube imports, and microphone or system-audio recordings all count). If you opt in once to the auto-summary feature, every transcript also receives an automatic on-device AI executive summary at no cost. The $39.99 one-time purchase removes the daily transcription cap and unlocks the rest of the AI Transcript Tools (regenerated summaries, titles, threads, and full AI chat). That is the entire monetization model.
Purchases
The $39.99 lifetime upgrade is sold and processed by Apple via the Mac App Store. We never see your name, email, payment method, or billing address. Apple tells the app whether you've purchased, and that's the only purchase information our app receives.
What we don't collect
- No account creation, no username, no email
- No analytics SDK that tracks you across the app
- No crash reporting that includes personal information
- No microphone or Screen Recording access unless you explicitly start a recording that needs it
- No data sold to data brokers, ever
Recording (optional)
The Record toolbar button offers three sources: Microphone (your voice), System Audio (whatever is playing on your Mac, for example a Zoom call), and Mic + System (both, mixed into one track). Microphone capture requires the Microphone permission. System Audio capture requires the Screen Recording permission; we use it only to capture audio, no screen contents are read or stored. Whichever source you pick, the recording is written to a .wav file on your Mac and transcribed locally. Recordings stay on your Mac so you do not lose them if anything goes wrong. You can delete an individual recording from its transcript, or clear every saved recording from Settings → Storage. Nothing is uploaded.
Children
Austin's Transcriptor carries a 4+ age rating in the Mac App Store because it contains no objectionable content, not because the app is designed or marketed for children. The app is not directed at children. We collect no personal information from any user, including children.
Sale or sharing of personal information
We do not sell or share personal information for advertising or for any other purpose. Because we don't collect personal information in the first place, there is nothing to sell or share. This includes any "sale" or "sharing" as defined by the California Consumer Privacy Act (CCPA) and similar state laws.
Your choices
- Unlock unlimited transcriptions and full AI Tools: hit the 5-per-day cap, or tap any AI card beyond the free auto-summary (titles, threads, ask anything, regenerate), to see the paywall. Complete the one-time $39.99 purchase, or choose Restore Purchases from the App menu to re-sync an existing purchase.
- Microphone: System Settings → Privacy & Security → Microphone → toggle Austin's Transcriptor on or off.
- Screen Recording (system audio): System Settings → Privacy & Security → Screen Recording → toggle Austin's Transcriptor on or off.
- Delete a transcript: right-click in the sidebar → Delete. Or move/empty the folder at
~/Library/Application Support/AustinsTranscriptor/to clear everything.
Changes
If this policy materially changes we'll update the effective date at the top and, where appropriate, surface the change in-app on the next launch.