MarketCAT
Speak instead of typing. The text pastes itself.
VoiceCat turns your voice into text. Hold a key, speak and let go, and the text appears wherever your cursor is: in an email, a chat, a document or your browser. Your recordings never leave your computer.
Four minutes
and you know it all.
You speak 2.4 times
faster than you type.
On a keyboard we type 52 words per minute on average. Normal dictation in Polish comes to 125. A 100-word email takes 48 seconds instead of almost two minutes.
For one email that's a small thing. With one hour of writing a day, it adds up to a day and a half of work each month.
| How much you write per day | Saved per week | Saved per month |
|---|---|---|
| 30 minutes | about 1.5 hours | about 6 hours |
| 1 hour | about 3 hours | about 12 hours |
| 2 hours | almost 6 hours | about 24 hours |
In practice: the program's author dictated over 104,000 words in 18 weeks. On a keyboard that would have taken about 20 hours longer.
How we calculated this: 52 words per minute is the average from a study by the University of Cambridge and Aalto University of over 168,000 people. 125 words per minute is the median of 1,183 real dictations in Polish by one person. A week is 5 working days and a month is 21. This is an estimate, not a guarantee: your pace may differ, and the time for any corrections depends on the text.
Hold the key.
Speak. Let go.
Click where you want the text
In any program, just as you would before typing.
Hold the key and speak
In your normal voice and your own words. The screen shows that recording is on.
Release the key
A moment later the whole text is pasted at your cursor. You can carry on speaking right away.
Works wherever
you write.
Email, chat, Word, browser. With VoiceCat you never copy anything over from a separate window: the text is pasted right at your cursor.
Works anywhere you can paste text.
Works even
without internet.
Speech is turned into text on your computer, so you can dictate on a train or without Wi-Fi too.
We don't send your recordings anywhere. You only need internet the first time you launch it.
No text
is ever lost.
The recording is saved to disk straight away, and every text stays in History.
If your cursor was in the wrong place, open History and copy the text with one click.
It transcribes
recordings and videos too.
Drop in a meeting recording, an interview or a video. You can also paste a link from YouTube, Instagram or TikTok. VoiceCat turns it into text you can copy and SRT subtitles that YouTube and any video editor accept.
From a link, it downloads only the audio, from public videos. Transcription happens on your computer, and you can keep dictating as usual in the meantime.
You don't need
ChatGPT.
Dictation works fully without any AI. If you want, you can connect your own account to fix punctuation and typos before pasting.
Only the dictated text then goes to the provider, and you pay them directly for usage. If the AI doesn't respond in time, the plain text is pasted.
OpenAI
ChatGPT models, via your API key.
Anthropic
Claude models, via your API key.
Claude Code
If you have it on your computer, you don't need a key.
Ollama
A model on your computer, no cloud.
You don't change
your habits.
No waiting
On an M1 iMac, a 41-second recording became text in under a second. On a Windows laptop with an Intel Core i5, a minute of speech takes about 4 seconds.
Your key
Any key or combination you choose. Hold it and speak, or toggle with a single press.
No minute limit
Everything happens on your computer, so you don't pay for dictation time. A forgotten recording stops on its own.
You pay once.
No subscription.
All packages include the same things. The only difference is the number of computers you can activate the key on.
Check before you buy: VoiceCat runs on Macs with an Apple processor (M1 or newer) and macOS 14 or newer, and on computers with Windows 10 or 11 (64-bit, Intel or AMD processor, 8 GB of memory). It doesn't work on Intel Macs for now. The app and the manual are available in English, German and Polish.
Every package gives you
- The VoiceCat app for Mac and Windows, one key for both
- License key, for life, no renewal
- Updates included in the license price
- A speech model that runs on your computer
- Dictation history you can paste from again
- Transcription of recordings, videos and links into text and SRT subtitles
- Optional connection to your own AI
After buying, you get the key and a download link by email. On first launch you paste the key into the app and you're done.
The most common
questions.
Will VoiceCat work on my Mac?
Click the Apple logo in the top left corner and choose “About This Mac”. If the processor says Apple M1 or newer and the system is macOS 14 or newer, the app will work. Macs with Intel processors aren't supported for now.
Will VoiceCat work on my Windows computer?
You need the 64-bit version of Windows 10 or 11, an Intel or AMD processor, 8 GB of memory and 3 GB of free disk space. On a slower computer, turning speech into text takes longer.
Do I need internet?
Only at the start: to activate the key and download the speech model, about 460 MB on Mac and 830 MB on Windows. After that, dictation works offline. When there's a connection, the app confirms the license once a day, and if there isn't, nothing gets blocked.
Do my recordings go anywhere?
No. Recordings and text stay on your computer. The license server only receives the key and a hash of the computer's ID. If you turn on AI correction yourself, the dictated text goes to your chosen provider, but not the recording.
Do I need ChatGPT or another AI?
No. The app works fully without AI and without any extra account. AI correction is an option you can turn on in settings.
How well does it understand English?
It handles normal speech well, but it isn't flawless. It most often gets company names, surnames and rare words wrong, so it's worth a quick look at the text before sending.
Which recordings and links can it transcribe?
Audio and video files such as MP4, MOV, MP3, M4A and WAV, and public videos from YouTube, Instagram and TikTok via a link. Downloading from a link needs the internet, and with the first link the app also fetches a download add-on, which takes about half a minute. It can't download private videos or ones that require signing in.
Is there a dictation limit?
There's no minute limit and no extra fees, because everything happens on your computer. So that a forgotten recording doesn't run all night, a single recording stops on its own after 5 minutes of silence or after 2 hours. The text then goes to History.
What if I change computers?
In the app's settings you disconnect the old computer, then enter the same key on the new one. If the old computer no longer works, write to kontakt@aicat.pl.
Is there a Windows version?
Yes. The same key works on Mac and Windows, and each computer takes up one seat on the key.
Windows shows a warning during installation. Is that normal?
Yes. VoiceCat is a new program and doesn't have a digital signature yet, so Windows shows the message “Windows protected your PC”. Click “More info”, then “Run anyway”.
How do updates work?
Once a day the app checks for a new version and shows it in the cat menu. You click, download and install, and your license, history and settings stay. Updates are included in the license price.
Speak instead of typing.
VoiceCat is an AiCat product, available in the MarketCAT store. You pay once, and your voice stays on your computer.