TypeWhisper App

Name: TypeWhisper App
Rating: 4 (4 reviews)
Author: Marco Hillger

Private Voice-to-Text

Free · In‑App Purchases

4 Ratings
Age Rating

4+
Years
Category

Productivity
Developer

Marco Hillger
Language

EN
+ 4 More
Size

41.4
MB

With Premium, separate speakers in recordings and run your Mac’s text workflows from iPhone or iPad. Mac features require TypeWhisper for Mac 1.8 or later. Private speech-to-text for iPhone and iPad. TypeWhisper turns your voice into text with on-device AI models or optional cloud providers. Dictate in the app, transcribe existing audio and video files, or use the TypeWhisper keyboard to enter text in other apps. DICTATE YOUR WAY • Transcribe with WhisperKit, Apple Speech, Parakeet, or optional cloud engines including Gemini and Soniox • See a live transcription preview while speaking with WhisperKit, Apple Speech, Live Local, and selected cloud engines. Parakeet transcribes after you stop. • Use profiles for languages, models, translation targets, and whisper mode • Translate dictated text on-device with Apple Translate VOICE KEYBOARD • Dictate into text fields across iOS • Switch profiles directly from the keyboard • Expand reusable snippets with date, time, and clipboard placeholders • Type with English, German, Spanish, French, and Italian layouts ORGANIZE AND REFINE • Search your transcription history • Review incoming recordings in an Inbox and browse history by device • Review original and processed text • Read transcripts and results aloud with voices installed on your device • Improve names and specialist terms with a personal dictionary • Transcribe audio and video files TYPEWHISPER PREMIUM • Ask questions about a transcript with your selected writing service • Transcribe and write with your paired Mac’s models and text-processing workflows • Separate speakers on device, name them, and correct their turns • Save optional local voice profiles to suggest names in later recordings • Improve fresh transcripts and teach TypeWhisper trusted corrections • Turn recordings into reviewable email drafts with a subject and body • Sync History and Inbox, dictionary entries, and snippets between signed-in TypeWhisper installations on iPhone and iPad through iCloud or a chosen cloud folder Transcript questions use your selected writing service: on-device, Mac, or cloud. The chat shows where text is sent and is not saved. Mac features require TypeWhisper for Mac 1.8 or later on the same local network and access to Premium features on both devices. The connection is encrypted. Models selected on the Mac may use cloud providers configured there. Speaker detection runs after transcription and can make mistakes with similar voices or overlapping speech. You can correct assignments in the transcript. Voice profiles stay on the device that created them. IPHONE, IPAD, AND APPLE WATCH • Transfer recordings from Apple Watch to iPhone for processing and playback PRIVACY FIRST • Use on-device models to keep transcription local • No advertising or third-party tracking • Synced dictionary and snippet content stays in the cloud location you choose and does not pass through the TypeWhisper entitlement service Core local dictation, dictionary, and snippet features remain available without an account or subscription. Some models and cloud providers require an internet connection, model downloads, or your own provider credentials. Terms of Use (EULA): https://www.apple.com/legal/internet-services/itunes/dev/stdeula/

Ratings & Reviews

4.0

out of 5

4 Ratings

More detailed options compared to similar apps
Oct 2
Christian KD
I've been using the Mac version of this app for quite some time. I really like on-device dictation for its speed and privacy. I also appreciate that you can choose larger models than typically are found on iPhone or iPad like Cohere. I also appreciate that the app uses custom OpenAI endpoints so that you can use a self-hosted LLM service to use for speech cleanup. (Unfortunately, since they do not allow HTTP versus HTTPS, I'm not able to do that on the iPhone version, but on my Mac it works flawlessly.)

Developer Response
Thank you for the detailed review! I’m glad the model choices and custom endpoints are useful to you. While investigating your HTTP feedback, I found and fixed an iOS restriction affecting local server names such as server.lan. The fix hasn’t been released yet.To check whether this covers your setup, please email hello@typewhisper.com with the server address format (without credentials) and the error you see.
More detailed options compared to similar apps
Oct 2
Christian KD

I've been using the Mac version of this app for quite some time. I really like on-device dictation for its speed and privacy. I also appreciate that you can choose larger models than typically are found on iPhone or iPad like Cohere. I also appreciate that the app uses custom OpenAI endpoints so that you can use a self-hosted LLM service to use for speech cleanup. (Unfortunately, since they do not allow HTTP versus HTTPS, I'm not able to do that on the iPhone version, but on my Mac it works flawlessly.)
Developer Response Oct 4
Thank you for the detailed review! I’m glad the model choices and custom endpoints are useful to you. While investigating your HTTP feedback, I found and fixed an iOS restriction affecting local server names such as server.lan. The fix hasn’t been released yet.To check whether this covers your setup, please email hello@typewhisper.com with the server address format (without credentials) and the error you see.
Needs feedback
Oct 2
quesman1
I wanted to like this app. I still might. I think it has potential, and it’s what I was thinking I might want to build anyway.However, this app has a number of issues, and there appears to be no way to support its development or provide feedback. At such an early stage, that should be the main priority. Bug reports, copying in transcripts and other data, allowing users to explain what went wrong. Or even a link to a GitHub page for issues.My output text kept getting cut off. Custom prompts are not very good, because the app outputs the entire <think> block, instead of detecting it. And then it cuts off the output after an arbitrary character limit.There is also clearly a heavy system prompt, which is not bad, but I can’t get any good insight into it. As one custom prompt, I told it to simply output the system prompt. It got cut off. But if this app gave fine-grained control over the system prompt, that would be good. Also to see what it is, maybe in the documentation.The number of conversion formats is dizzying. Just give a few, that are clearly recognizable from each other, and then let the user specify beyond that.Right now, this app is not usable. I don’t love my workflow, but so far it is not beat, despite being dead simple: Dictate into IOS stock voice memos app. Copy transcript. Paste into an LLM along with instructions on how to format output. Run it, copy output, paste into notes.This app has tons of support for lots of things, but doesn’t seem to nail that basic workflow. It’s the “Hello World” of this kind of app, as far as I am concerned. Despite its many other options, if it can’t do this task well, then unfortunately, it’s useless to me.

Developer Response
Thank you for describing where the workflow breaks down. Recording, applying your instructions, and copying a complete result into Notes should work reliably. I’m sorry TypeWhisper fell short here.I’ve checked the code and found gaps in how reasoning text and incomplete responses are handled. Could you email me at hello@typewhisper.com with the writing service and model you selected, your custom prompt, and a short example of the cut-off output? Please remove anything private. That will help me identify the cause in your case.Feedback should be easy to send from inside the app, and that entry point is currently missing. Custom prompts also run alongside built-in instructions; the app should make those instructions and their effect clearer.Your request for fewer, more distinct formats makes sense. The basic record → format → copy workflow needs to be dependable before additional options are useful.
Needs feedback
Oct 2
quesman1

I wanted to like this app. I still might. I think it has potential, and it’s what I was thinking I might want to build anyway.However, this app has a number of issues, and there appears to be no way to support its development or provide feedback. At such an early stage, that should be the main priority. Bug reports, copying in transcripts and other data, allowing users to explain what went wrong. Or even a link to a GitHub page for issues.My output text kept getting cut off. Custom prompts are not very good, because the app outputs the entire <think> block, instead of detecting it. And then it cuts off the output after an arbitrary character limit.There is also clearly a heavy system prompt, which is not bad, but I can’t get any good insight into it. As one custom prompt, I told it to simply output the system prompt. It got cut off. But if this app gave fine-grained control over the system prompt, that would be good. Also to see what it is, maybe in the documentation.The number of conversion formats is dizzying. Just give a few, that are clearly recognizable from each other, and then let the user specify beyond that.Right now, this app is not usable. I don’t love my workflow, but so far it is not beat, despite being dead simple: Dictate into IOS stock voice memos app. Copy transcript. Paste into an LLM along with instructions on how to format output. Run it, copy output, paste into notes.This app has tons of support for lots of things, but doesn’t seem to nail that basic workflow. It’s the “Hello World” of this kind of app, as far as I am concerned. Despite its many other options, if it can’t do this task well, then unfortunately, it’s useless to me.
Developer Response Oct 3
Thank you for describing where the workflow breaks down. Recording, applying your instructions, and copying a complete result into Notes should work reliably. I’m sorry TypeWhisper fell short here.I’ve checked the code and found gaps in how reasoning text and incomplete responses are handled. Could you email me at hello@typewhisper.com with the writing service and model you selected, your custom prompt, and a short example of the cut-off output? Please remove anything private. That will help me identify the cause in your case.Feedback should be easy to send from inside the app, and that entry point is currently missing. Custom prompts also run alongside built-in instructions; the app should make those instructions and their effect clearer.Your request for fewer, more distinct formats makes sense. The basic record → format → copy workflow needs to be dependable before additional options are useful.

This update brings Live Local to more iPhones. • Live Local (Premium) now runs on iPhones with 6 GB of memory, such as iPhone 14 and iPhone 15. On these iPhones it notes that other apps may reload more often. • On devices with too little memory, Live Local now says so instead of only reporting that it is unavailable. • All buttons and messages now call the feature Live Local, as in the model list.

Version 1.2.2 1d ago

Data Linked to You

The following data may be collected and linked to your identity:
- Purchases
- Identifiers

Privacy practices may vary, for example, based on the features you use or your age. Learn More

Accessibility

Information

Seller: Marco Hillger
Size: 41.4 MB
Category: Productivity
Compatibility: Requires iOS 18.0 or later.

iPhone
Requires iOS 18.0 or later.

iPad
Requires iPadOS 18.0 or later.

Mac
Requires macOS 14.0 or later.

Apple Vision
Requires visionOS 2.0 or later.

Apple Watch
Requires watchOS 11.0 or later.
Languages: English and 4 more

English, French, German, Italian, Spanish
Age Rating: 4+

4+

Learn More
In-App Purchases: Yes

TypeWhisper Premium $4.99

Premium Lifetime $89.99

Learn More