Support

No. Recordings, transcripts, and summaries stay on your iPhone. The only network request the app ever makes is downloading the Parakeet model weights from huggingface.co/FluidInference/parakeet-tdt-0.6b-v3-coreml on first use, and Apple Translation if you tap "Translate Transcript". See the privacy policy for details.

Two usual causes: (1) the recording is too quiet — try closer to the mic; (2) the recording is in a language Parakeet doesn't cover (see the language list below) — the app automatically retries with Apple Speech in that case, but you can also set Language to "Auto-detect" first.

Yes. It's a single 0.6B-parameter model that runs comfortably on the Apple Neural Engine from iPhone 12 onward; a newer iPhone just transcribes faster. Apple Speech (Free tier) works on all iPhones with iOS 17+.

Only for the AI Summary, AI Q&A, and Auto-tag features (the Pro+ tier). The base transcription works on every iPhone running iOS 17+. Apple Intelligence requires iPhone 15 Pro / 16 / 17 and iOS 26.

Open Settings → tap your name → Subscriptions → Echograph Pro+ → Cancel Subscription. Your subscription stays active until the end of the billing period.

Pro covers everything most people need — Parakeet, custom vocabulary, exports — for a single $19.99 payment. Pro+ adds AI features that depend on continuously-updated Apple Intelligence models, which is why it's a yearly subscription. You can stay on Pro forever and never pay again.

Yes. The watchOS companion app records audio standalone (great for one-handed capture), then transfers the file to your iPhone over WatchConnectivity when you stop. Transcription happens on the phone.

Yes. Tap the ••• menu in any recording → Export → choose Markdown for Obsidian/Bear, Word/PDF for traditional notes, SRT/VTT for video editors. Or use Share to AirDrop, Mail, or any iOS share extension.