Say it. Franz types it.
Hold one key, talk, let go. The words land in whatever you were writing in.
Voice
Works in every service you've added.
The transcript appears while you speak, and the text drops in the moment you let go of the key. Nothing gets sent anywhere: the recognition runs on your Mac.
It doesn't care which service you're in. WhatsApp, Slack, a web app you added yourself, the reply box on a support ticket. If you can type there, you can talk there instead. Franz doesn't ask the service for permission, and the service never notices the difference. You pick the key in Settings, under Voice, and you can change it whenever you like.
How it works
- 01
Hold your key. A pill appears above whatever you're writing in.
- 02
Talk. The words show up as you go, so you can watch it keep up.
- 03
Let go. The text lands in the field, and you edit it like anything you typed yourself.
Why it reads better
Typed replies come out colder than you mean them.
You know the answer you'd have given out loud. Warm, a bit rambling, with the apology in the middle where it belongs. Then you type it and it turns into four words and a full stop. Nobody reads that as friendly. You were just tired of typing.
Talking gets the first version out the way you'd actually say it. You still edit it and you still cut half of it. But you're editing something warm instead of building something short. Franz sits at both ends of that trade. It already transcribes the voice notes that land in your chats. Now it works the other way round too.
On your Mac
Your voice doesn't leave the machine.
Dictation runs on Apple's on-device speech recognition, through a small helper that ships inside Franz. There's no server in this path at all, and no cloud mode to switch on. We never see a word of it, because none of it goes anywhere.
One thing worth keeping apart. Transcribing the voice notes other people send you is a different feature, and that one has a cloud option you can pick. Dictation doesn't. It's on-device or it's off.
Languages
You don't have to switch to English.
Dictation handles German, English, Spanish, French, Italian, Japanese, Korean, Portuguese, Cantonese and Chinese. Apple's recognizer is what sets that list, so it can grow with a macOS update. You pick a main language in Settings, and a second one if you go back and forth. Franz downloads a language the first time you use it, and macOS keeps it from there.
- German
- English
- Spanish
- French
- Italian
- Japanese
- Korean
- Portuguese
- Cantonese
- Chinese
Everywhere you write
Including the long emails.
Anything with a text field takes it. Your messengers, the web apps you added yourself, and Franz Mail, where the replies get long. That last one is where dictation earns its keep, because a proper email is where typing hurts most. We haven't found another all-in-one messenger you can talk into. A few of them transcribe the voice notes you receive, which is the opposite job.
Questions
The things people ask first.
No. Dictation uses Apple's on-device speech recognition, so it needs macOS 26 or later on an Apple Silicon Mac.
Apple Silicon Macs on macOS 26 or later. Intel Macs can't: Apple's recognizer needs the Neural Engine that Apple Silicon has.
Franz Pro. Pro Cloud and Lifetime include it too, and it's part of the trial. You get it whether you downloaded Franz from us or from the Mac App Store.
No. Dictation has no cloud path at all. Transcribing the voice notes other people send you is a separate feature, and that one does offer a cloud option you choose yourself.
German, English, Spanish, French, Italian, Japanese, Korean, Portuguese, Cantonese and Chinese. You set a main language in Settings and can add a second.
Yes. Dictation types into whichever field you're in, so a web app you added yourself behaves the same as a service Franz ships.
Ready to take control?
Downloaded over 1 million times. Franz brings messaging, email, and client work into one workspace.