Support

Bug reports, feature requests, and feedback all go to one inbox.

Contact

Email austindmunday@gmail.com. Responses usually within a couple of days.

Common questions

How accurate is the transcription?

It depends on the model and the audio. The Tiny model (bundled, runs instantly) is rough but usable for voice memos. Small and Medium are good for podcasts and meetings. Large-v3 is the best Whisper can do — it's slower and uses more memory, but the accuracy is excellent for difficult audio (accents, background noise, technical jargon).

How long does transcription take?

Roughly: Tiny is 15× realtime, Base is 8× realtime, Small is 4× realtime, Medium is 2× realtime, Large-v3 is around realtime — all on Apple Silicon. So a one-hour file takes ~4 minutes with Tiny, ~15 minutes with Small, and ~60 minutes with Large-v3. Times are estimates and depend on your specific Mac.

Why does the first launch download a file?

It doesn't, by default. The Tiny model is bundled with the app so transcription works immediately. The first time you pick Small, Medium, or Large, the app downloads that model from Hugging Face (between 244 MB and 1.55 GB). Subsequent transcriptions reuse the cached model.

The speaker labels are wrong

Diarization is harder than transcription, and it's noisier when speakers talk over each other or when a clip is short (under ~20 seconds of audio per speaker). You can right-click any segment to reassign it to the correct speaker, and you can rename "Speaker 1 / Speaker 2 / …" to actual names — the names save with the transcript.

I bought the lifetime upgrade but I still see ads

Tap "Restore Purchase" in the paywall sheet. This re-syncs your purchase state from Apple. If that doesn't fix it, quit and relaunch the app. Make sure you're signed into the Mac App Store with the same Apple ID you used to buy the upgrade.

Can I transcribe Zoom or system audio?

Yes. Click the "Record System Audio" button in the toolbar. macOS will ask for Screen Recording permission the first time — grant it in System Settings → Privacy & Security → Screen Recording. The captured audio is processed entirely on your Mac.

Where are my transcripts stored?

~/Library/Application Support/AustinsTranscriptor/transcripts/ as JSON files. You can back them up, sync them via iCloud Drive, or delete them.

Does my audio ever leave my Mac?

No. The only network traffic the app makes is downloading model files from Hugging Face on first use, and loading the sidebar banner ad if you're on the free tier. Your audio and transcripts are local-only. See the Privacy Policy for the full breakdown.

Refunds

Apple processes all refund requests for Mac App Store purchases. File a request at reportaproblem.apple.com.

I get an error: "Transcription failed"

Most often this is a memory issue with the Large model on a Mac with less than 16 GB of RAM. Try the Medium or Small model — they're nearly as accurate and use much less memory. If a smaller model also fails, email me with the audio format and your Mac model.

Cancel a subscription

There is no subscription. The $39.99 ad-free upgrade is a one-time purchase. There's nothing to cancel.