Voice to Text for MacOS
Dictation That Gets Smarter
$20 Lifetime License 100% On-Device
Your voice never leaves your Mac
No Subscription
$20 once — yours forever
Dual Engine
Two Apple speech engines, one result
Everything you need from dictation
Built on Apple's native speech engines. No cloud. No compromise.
Dual-Engine Recognition
DictationTranscriber handles natural flow and punctuation. SFSpeechRecognizer boosts your custom vocabulary. Both run simultaneously on the same audio for best-of-both accuracy.
Custom Vocabulary
500 terms with auto-learn from corrections. Technical jargon, brand names, medical terms — Hush Touch recognizes what you actually say. Per-app profiles adapt automatically.
Smart Text Processing
Filler word removal, auto-correction ("actually" detection), voice snippets for repeated phrases, and numbered list formatting — all processed locally in real time.
Context Modes
Switch between General, Email, Code, and Notes modes. Each profile tunes text processing — formal punctuation for email, camelCase preservation for code, bullet points for notes.
Adaptive Learning
Hush Touch detects when you correct a transcription and auto-adds the right word to your vocabulary. Frequency-weighted language models with per-app profile boosting.
100% On-Device Privacy
Zero cloud dependency. Zero data collection. Your voice is processed entirely by Apple's native speech engines on your Mac. "Data Not Collected" App Store privacy label.
Two engines. One perfect result.
See how Hush Touch merges Apple's native speech engines in real time.
Your voice
"Deploy the app to Kubernetes and use kubectl to manage pods."
DictationTranscriber
Natural flow + punctuation
SFSpeechRecognizer
Custom vocabulary spotter
Merged result
"Deploy the app to Kubernetes and use kubectl to manage pods."
See how Hush Touch compares
Premium features. No subscription tax.
| Feature | Hush Touch $20 once | Wispr Flow $12/mo | Superwhisper $8.49/mo |
|---|---|---|---|
| Price | $20 once | $144/yr | $102/yr |
| Privacy | 100% on-device | Cloud-only | Hybrid |
| Custom vocabulary | 500 terms + auto-learn | Personal dictionary | Custom vocabulary |
| Auto-punctuation | Native engine | Cloud AI | Cloud AI |
| Filler removal | On-device | Cloud AI | Cloud AI |
| Context modes | 4 modes | Tone/style modes | Custom AI modes |
| Adaptive learning | Correction detection + frequency weighting | Cloud learning | No |
| Per-app profiles | 10 profiles with CLM boosting | No | App-aware context |
| Voice snippets | 50 mappings | No | No |
| Offline | Always | Never | Optional |
| Dual engine | DictationTranscriber + SFSpeechRecognizer | Single cloud engine | Multiple models |
Your voice never leaves your Mac
Hush Touch uses Apple's on-device speech engines exclusively. No cloud. No exceptions.
- All speech processing happens on-device via Apple's native engines
- Zero network requests during dictation — verified by any network monitor
- No analytics, no telemetry, no user tracking
- "Data Not Collected" App Store privacy nutrition label
- Works fully offline — no internet required, ever
Lifetime license
One-time purchase. No subscription.
- Dual-engine speech recognition
- Custom vocabulary (500 terms)
- Adaptive learning from corrections
- Smart text processing pipeline
- 4 context modes (General/Email/Code/Notes)
- Per-app vocabulary profiles
- Voice snippets (50 mappings)
- Free lifetime updates
- 100% on-device — no cloud ever