Start here
Setup, proof, and the links people actually need.
This page should get someone from “what is this?” to “I can use it” without dumping them into random bullets. Start with setup, then use the proof and comparison links if you need them.
App Store rating: 4.9 / 5. See reviews.
Quick answers
Speech to text on Mac
The practical orientation page if you are starting from Apple Dictation and working outward.
How to use speech to text on Mac
Short version if you just want the workflow and not the whole category page.
Mac Dictation commands
Punctuation, formatting, and line-break commands in one place.
Setup
Working in 30 seconds.
Step 1
Download from the Mac App Store and open Voice Type.
Step 2
Choose your preferred model (optimized for Apple silicon, works on Intel too).
Step 3
Press and hold your hotkey (default "⌥ Space"), speak, release to finalise.
Why it feels different
The useful product facts, without the chest-thumping.
Hotkey voice typing anywhere
Trigger dictation from any app with a single hold-to-dictate shortcut. When you release, Voice Type finalises the last few seconds and inserts the text where your cursor lives.
Accuracy you can feel
Voice Type uses beam search decoding rather than greedy decoding, exploring multiple transcriptions simultaneously for higher accuracy on complex phrases. We condition audio before recognition so technical terms and product names transcribe correctly.
Custom vocabulary uses prompt conditioning following Whisper best practices. See examples on the Accuracy page.
Dragon-level punctuation
Voice Type has near-parity with Dragon Dictate for spoken punctuation. Say "period", "comma", "question mark", "new line", and "new paragraph" naturally while speaking. No weird syntax, no ritual pauses.
Sub-2-second finalization
Voice Type finalizes in under 2 seconds regardless of how long you dictate. The streaming architecture processes audio in ~30-second windows as you speak, so when you release the hotkey, only the last segment needs finalizing.
Processing runs entirely on-device using Apple's Core ML and Metal for GPU acceleration. Try the latency demo.
Troubleshooting
Fix the obvious blockers first.
Compare options
Use the comparisons when you are evaluating, not when you are just trying to get set up.
FAQ
The questions people actually ask before they install.
- How accurate is Voice Type?
- We improve clarity before recognition and optimise for everyday technical terms. See live examples on the Accuracy page.
- How do I turn on voice typing on a Mac?
- → System Settings → Keyboard → Dictation → On. Voice Type adds a global hotkey so you can talk in any app.
- Is Voice Type offline?
- Yes. Processing runs on your Mac. Accuracy, speed, and privacy come first; offline processing keeps latency steady.
- Why does offline dictation feel faster?
- No network round-trips. When you release the hotkey, we only finalise the last few seconds before inserting text.
- Does Voice Type require an account?
- No. Purchase once on the Mac App Store after the 7-day trial.
- What Mac models are supported?
- All Macs that support Apple’s on-device speech frameworks, including Apple silicon and Intel models.
References
Further reading, if you want receipts.
- OpenAI Whisper (GitHub)
The open-source speech recognition model architecture
- Robust Speech Recognition via Large-Scale Weak Supervision
Original Whisper research paper (OpenAI, 2022)
- RNNoise (GitHub)
Recurrent neural network for audio noise suppression
- RNNoise: Learning Noise Suppression
Technical demo and explanation by Jean-Marc Valin
- Apple Core ML Documentation
On-device machine learning framework
- Apple: Dictate messages and documents on Mac
Official macOS dictation guide
- Hacker News: Whisper large-v2 discussion
Community discussion on Whisper accuracy improvements