Comparison, as of Oct 16, 2026
Superwhisper Alternative for iPhone: SpeechDone Compared
According to the vendor’s documentation, Superwhisper can run entirely on the device and is available for macOS, Windows, iOS, and Android, which is its strength. SpeechDone fits better if you dictate on iPhone, iPad, or Mac, accept recognition through servers, and want a subscription, a dictionary, and a history in your iCloud in return.
SpeechDone and Superwhisper side by side
| SpeechDoneWaitlist | SuperwhisperAccording to the vendor’s documentation | |
|---|---|---|
| Availability | Waitlist, not yet in the App Store | iOS app availableSource 2 |
| Devices | iPhone, iPad, and Mac (macOS 15 or later) | macOS, Windows, iOS, and Android. On iOS, iPhone only, iPad is not supportedSource 2Source 6 |
| Keyboard on iPhone | A dictation keyboard of its own in every app | A keyboard of its own, needs “Full Access”Source 2 |
| Free tier | 1,500 words a week | Unlimited dictation with Whisper models on the device, up to 2 modes without AI processing. New users get 3,000 words of access to all Pro featuresSource 1 |
| Pro | $4.99 a month or $49.99 a year, no weekly limit. Subscription through the App Store | $8.49 a month, $84.99 a year, or $249.99 once. The vendor says the App Store price is higherSource 1 |
| Where speech is processed | Our server is hosted by Hetzner in Germany. Recognition runs only on EU servers (Deepgram and Mistral AI) | You choose the model: on the device or cloud models. The documentation does not say where the cloud models runSource 5 |
| Dictionary and sync | Dictionary and text snippets are the same on all your devices (with sign-in) | Custom vocabulary available. On iOS, modes and vocabulary are stored per device and not synced from the Mac, according to the documentation (cloud sync is planned)Source 2 |
| DPA | DPA under Art. 28 GDPR, which you sign with us in text form. Plus a confidentiality obligation under § 203 of the German Criminal Code and a public list of subprocessors | DPA and BAA can be signed online. Subprocessors are listed in the vendor’s Trust CenterSource 4 |
As of Oct 16, 2026, per the vendor’s documentation and privacy page (sources below). SpeechDone prices include VAT.
In short
Superwhisper and SpeechDone take different routes. Superwhisper lets you pick models, including ones that run on the device. According to the vendor’s documentation, with a purely on-device setup neither audio nor text leaves the device. SpeechDone cannot promise that, because our recognition runs through servers.
Everything we say about Superwhisper comes from the vendor’s public documentation and privacy pages, as of Oct 16, 2026. Prices are in US dollars, as given there. About SpeechDone we only say what we can back up ourselves.
What Superwhisper does well
If no word should leave your device, Superwhisper is the better fit. You can use models such as Whisper Large v3 Turbo or Parakeet on the device, and the free tier allows unlimited dictation with the on-device Whisper models. If you would rather pay once than subscribe, there is a $249.99 license, and one Pro license covers all your devices, according to the vendor.
It also reaches further: macOS, Windows, iOS, and Android. Cloud models are optional, and the documentation names Anthropic, OpenAI, Deepgram, and Groq among the providers. According to the vendor they run under API contracts with zero data retention and no use for training (this applies to API use only).
What SpeechDone does differently
SpeechDone is built for iPhone, iPad, and Mac: a keyboard in every app on iPhone and iPad, hold fn and speak on the Mac (macOS 15 or later), plus a widget, a Live Activity, the share menu, and a history in your iCloud. You see the text while you speak, and when you stop, the finished text is in the field. Two speech recognizers check each other: Deepgram writes live, Mistral recognizes the recording again in sections, and that text is what gets inserted.
You do not have to deal with models. You also get a dictionary, text snippets on all your devices (with sign-in), and in the keyboard you can translate, ask questions, and rewrite by voice (Pro). Past the weekly limit, SpeechDone keeps writing, just without cleanup.
For context: Deepgram (USA, processing on EU servers) and Mistral AI (Paris) receive your recording for recognition. Neither uses it for training: we opt out with Deepgram on every request, and with Mistral it is switched off in our account. On our server, the recording is not stored.
When SpeechDone fits better
You do not want to set anything up
You do not have to pick models, SpeechDone takes recognition off your hands.
You use iPhone, iPad, and Mac
SpeechDone runs on iPad too, which the Superwhisper documentation says is not supported.
You need a server in Germany and a DPA in text form
Our server is hosted by Hetzner in Germany, and recognition runs only on EU servers. The list of subprocessors is public.
When Superwhisper fits better
Nothing should leave the device
With on-device models, neither audio nor text leaves the device, according to the vendor.
You also work on Windows or Android
One Pro license covers all your devices, according to the vendor.
You want to pay once
Besides the subscription there is a $249.99 license.
Frequently asked questions
Does Superwhisper run entirely on the device?
According to the vendor’s documentation, yes, if you use only on-device models. Cloud models are an additional option. SpeechDone processes speech through servers.
How much does Superwhisper cost?
According to the vendor, $8.49 a month, $84.99 a year, or $249.99 once, plus a free tier. SpeechDone Pro is $4.99 a month or $49.99 a year.
Is there a Superwhisper app for iPad?
According to the vendor’s documentation, the iOS app supports iPhone only. SpeechDone runs on iPhone, iPad and Mac.
Can I sign a DPA with Superwhisper?
According to the vendor, a DPA and a BAA can be signed online. With SpeechDone you sign the DPA with us in text form.
Where do I get SpeechDone?
SpeechDone is not in the App Store yet. Join the waitlist and we will email you once when it is.
Sources
- 1Superwhisper: Plans, retrieved October 16, 2026
- 2Superwhisper: iOS, retrieved October 16, 2026
- 3Superwhisper: Sensitive data, retrieved October 16, 2026
- 4Superwhisper: Compliance, retrieved October 16, 2026
- 5Superwhisper: Voice models, retrieved October 16, 2026
- 6Superwhisper: documentation index with platform list, retrieved October 16, 2026