Compare
VV vs Superwhisper: two local-first apps, different defaults
This is not a privacy argument. Both can recognise speech on your Mac. The real difference is how much of the machinery you want to configure, and what happens by default when you do not.
Last updated
Why this comparison is unusual
Most dictation comparisons come down to where the audio is processed. This one does not. Superwhisper says it works offline, so you can transcribe anytime
, and it can run local models on Apple Silicon. So can VV. On the question that usually decides everything, the two agree.
What is left is a difference of philosophy, and it is a real one. Superwhisper is a toolkit: many models, many modes, bring your own API keys, three platforms. VV is an appliance: one shortcut, sensible defaults, and a deliberately small number of things to get wrong.
| Superwhisper | VV | |
|---|---|---|
| Local recognition | Yes — best on Apple Silicon, per their FAQ | Yes, and it is the default |
| Cloud models | Yes, including your own API keys on Pro | Only for an optional rewrite you configure |
| Platforms | Mac, Windows, iOS | Mac (Apple Silicon) and Windows 10/11 (x64); one licence covers both; no phone app |
| Intel Macs | Supported, though their FAQ says Intel "works best with Cloud models" | Not supported — Apple Silicon only |
| Configuration | Extensive: models, modes, prompts | Deliberately narrow |
| Free tier | Yes | 1,000 words a week, no sign-in |
| Pricing shape | Free, Pro monthly or yearly, one-off Lifetime, Enterprise | Free, then Pro £9.99/month or £99/year, with Vaitly Meetings and AI editing |
| Refund window | 30 days, per their site | 14 days, no questions asked |
Superwhisper details read on superwhisper.com on 11 September 2026. We have deliberately not quoted their Pro price here, because we could not read it cleanly from their page on the day of checking — see their pricing page for the current figure.
Configurable versus opinionated
Superwhisper's strength is that almost everything is adjustable. You can pick the speech model, define modes for different kinds of writing, attach cloud AI models with your own keys, and shape what comes out. If you like that kind of control, it is genuinely excellent and you will get more out of it than you will out of us.
VV takes the other view: that most people want dictation the way they want a light switch. There is one shortcut. The speech model is chosen for your computer. Recognition is local, and that is not a mode you select. The settings that exist are the ones that change what you get — retention, vocabulary, snippets, a rewrite step — rather than the ones that change how it works underneath.
What VV does that Superwhisper does not
- Vocabulary with a spoken form. You teach VV not just the spelling of a name but how you say it, so it biases to your spelling when it hears that sound. This is the single biggest accuracy change for most people, and it is why a dictation app that knows your colleagues beats one that knows the dictionary.
- Snippets. A short spoken trigger expands to a whole block — a sign-off, a postal address, a paragraph you send weekly. See how to build voice snippets.
- Meetings in the same app. Pro includes Vaitly Meetings (beta), which records calls in the background with no meeting bot joining, then writes minutes, decisions and actions.
- No sign-in to dictate at all. Free needs no account, no email, no card.
What Superwhisper does that VV does not
- iOS. Superwhisper lists an iPhone app; VV has no phone app. VV runs on Mac and Windows, and one licence covers both, but if you dictate on a phone this is the honest deciding factor.
- Intel Mac support. VV requires Apple Silicon (M1 or later). If you are on an Intel Mac, VV is not an option and Superwhisper is.
- Bring your own API keys. If you already pay for a frontier model and want it driving your rewrites, Superwhisper's Pro tier supports that directly.
- Per-mode configuration. Different models and prompts for different kinds of writing, in a way VV deliberately does not expose.
The Apple Silicon question
Superwhisper's FAQ is refreshingly direct about this: Intel Macs work best with Cloud models. Offline models only run really well on Apple Silicon macs.
That is true of local speech recognition generally, not just of their product.
VV took the same fact to a different conclusion and simply requires Apple Silicon (M1 or later) on macOS 14 Sonoma or later. It means we never ship a configuration that is quietly bad, and it means Intel Mac owners cannot use us at all. That is a real cost and we would rather state it here than in a support reply. Read more in why Apple Silicon is good for local speech AI.
How to choose in fifteen minutes
- Check the hardware constraint first Intel Mac, or an iPhone in the mix? Take Superwhisper. The rest of this list will not change that answer. A 64-bit Windows PC alongside the Mac no longer decides it: VV runs on both under one licence.
- Install both free tiers Give them different shortcuts so they are not fighting for the same key.
- Dictate one real paragraph into each From your work, with your jargon in it. Not a test sentence.
- Notice what you did next Did you reach for settings, or did you carry on writing? That reaction is the answer to the toolkit-versus-appliance question, and it is more reliable than any opinion you hold about yourself.
- Teach each one five words The names you use daily. Whichever one made that easy is the one you will still be using in a month.
Where these facts came from
Anything stated about another product was read on that product's own site or documentation on the date shown, and is re-checked at least quarterly. If something below has changed, tell us and we will correct it.
- Superwhisper says it "works offline, so you can transcribe anytime," and its FAQ adds that "Intel Macs work best with Cloud models. Offline models only run really well on Apple Silicon macs." source, checked .
- Superwhisper lists Mac, Windows and iOS. source, checked .
- Superwhisper publishes a Free tier, a Pro plan billed monthly or yearly, a one-off Lifetime licence and custom Enterprise pricing, with a 30-day refund on all plans and a student discount. source, checked .
- Superwhisper lets Pro users bring their own API keys and use cloud AI models alongside local ones. source, checked .
Questions
Do both really work offline?
Both can. Superwhisper says it works offline and notes that offline models run best on Apple Silicon. VV runs its standard recognition on-device by default, on an Apple Silicon Mac or a Windows PC, and needs the internet only for the first-run model download. The way to be sure with either is the aeroplane-mode test described in our guide to on-device dictation.
Is one more accurate?
We have not run a benchmark we would be willing to publish, and we would treat anyone else's marketing benchmark with suspicion. Both can run strong local models. Accuracy in real use is dominated by your microphone, your room and whether the app knows your vocabulary — which is why we would rather you tested both on your own voice for ten minutes.
Can I run both?
Yes. Different shortcuts, and they will not interfere. Some people use one for quick dictation and the other for longer, more configured work.
Which is better value?
Depends entirely on how long you keep it and how many devices you need. Superwhisper publishes a Lifetime option and a 30-day refund; VV has a permanent Free plan and Pro at £9.99 a month or £99 a year, with a 14-day refund, and a paid plan covers 5 devices in any mix of Mac and Windows. If you need an iPhone app, Superwhisper is better value at any price, because VV does not have one.