VoiceInk real-time transcription: which models stream, and how to turn it on
By Ryleigh Newman · Published
VoiceInk can show your words as you speak, but only in a Mode whose model supports streaming. The Parakeet models do, and so do the Nemotron models and several cloud providers. Local Whisper, Apple Speech and Cohere Transcribe do not. You turn it on with the Real-time switch in a Mode's Transcription section.
I built YapToText, a different dictation app for the Mac, so read what I say about VoiceInk with that in mind. Every fact about VoiceInk here is from its documentation, read on October 9, 2026. The page is Real-time Transcription.
This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.
This guide in other languages: العربية, Čeština, Dansk, Deutsch, Ελληνικά, Español, Suomi, Français, עברית, हिन्दी, Magyar, Bahasa Indonesia, Italiano, 日本語, 한국어, Norsk, Nederlands, Polski, Português, Română, Русский, Slovenčina, Svenska, ไทย, Türkçe, Українська, Tiếng Việt, 简体中文, 繁體中文
Turn it on
- Open Modes and edit the Mode you dictate with.
- In Transcription, choose a model that supports streaming.
- Turn on Real-time.
- To see the partial text, check Settings, Interface, Live Text Display.
Hiding the live text display does not turn real-time processing off in the Mode. Both recorder styles, Notch and Mini, can show it.
Which models stream
From VoiceInk's docs page Real-time Transcription, read October 9, 2026. The page lists a few more cloud models.
| Models | Real-time |
|---|---|
| Parakeet V2, V3, Ultra and Unified | Optional |
| Nemotron Latin and Nemotron Multilingual | Required for live recording |
| Deepgram Nova 3 and Nova 3 Medical, ElevenLabs Scribe V2, Mistral Voxtral, Soniox V5, Speechmatics, AssemblyAI Universal-3.5 Pro | Optional, in the cloud |
| Cartesia Ink 2 | Required. Streaming only, and English only |
| Local Whisper, Apple Speech, Cohere Transcribe, SenseVoice | No streaming option |
What real-time changes, and what it does not
- Partial text can change. The words on screen may be revised as you keep speaking.
- The final text is still processed. Formatting, word replacements and optional AI enhancement run after recording ends.
- Cloud streaming sends audio as you speak. It needs a working connection for the whole recording.
- Files are different. Imported files use batch processing, and a streaming-only model such as Cartesia Ink 2 cannot process them.
The VoiceInk model guide covers which model to pick, and the privacy page covers what a cloud model receives.
Live text in YapToText
YapToText shows a live transcription preview in its panel while you talk, on by default: your voice as a wave, and your words as they form. The finished text goes into your app when you stop. A separate setting, Type text live as you speak, types the words into the app while you talk. It is off by default and experimental.
It does this with Whisper on the Mac, and the preview costs extra battery, so you can turn it off. VoiceInk's route to live text is a streaming model such as Parakeet. Parakeet vs Whisper explains the two model families, and the Live Captions guide covers Apple's live text for anything you hear.
This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.
The short version
- Switch. Real-time, in a Mode's Transcription section.
- Local streaming. Parakeet and Nemotron.
- No streaming. Local Whisper, Apple Speech and Cohere Transcribe.
- Display. Settings, Interface, Live Text Display.
- Files. Batch only.
If VoiceInk changes which models stream, open an issue on GitHub and I will update this page. Every one gets read.
This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.
Related
Sources
- https://tryvoiceink.com/docs/realtime-transcription (checked 2026-10-09)
- https://tryvoiceink.com/docs/local-models (checked 2026-10-09)
- https://tryvoiceink.com/docs/transcribe-audio-files (checked 2026-10-09)
- https://github.com/ryleighnewman/YapToText (checked 2026-10-10)
- https://apps.apple.com/us/app/yaptotext/id6786382289 (checked 2026-10-10)
Questions
Does VoiceInk have real-time transcription?
Yes, in Modes whose model supports streaming. Turn on Real-time in the Mode's Transcription section.
Which VoiceInk models support streaming?
Locally, the Parakeet and Nemotron models. Several cloud models do too. Local Whisper, Apple Speech and Cohere Transcribe do not.
Why is there no live text with Whisper in VoiceInk?
Its docs list local Whisper among the models that do not provide the streaming option.
How do I hide the live text in VoiceInk?
Use Settings, Interface, Live Text Display. Hiding it does not turn off real-time processing in the Mode.
Does VoiceInk real-time transcription work offline?
With a local streaming model such as Parakeet, yes. Cloud streaming needs a connection for the whole recording.