Say it once,
it lands everywhere.
Talk, and clean text appears where you're writing. Say what needs doing, and it becomes tasks in the tools you already use. Record a meeting and get the note. Speech never leaves your device unless you say so.
macOS 14 or later, Apple silicon. Free. No account needed.
Five things to do with your voice, and one it does on its own
Hold a key on the Mac, or tap Dictate on the iPhone keyboard. What happens to the words depends on the mode: on the Mac it follows the app you're in, and on the iPhone keyboard you pick the tone.
Dictate
Hold, speak, release. Fillers and false starts are gone, sentences are fixed, and the tone matches where you're writing. The Mac starts with a set you can change: casual in Slack, formal in Mail, technical in Xcode and VS Code, verbatim in Terminal and iTerm.
"um so can we uh move the call to thursday, actually no friday"Edit
Select some text, hold the edit key, say how to change it. The edited version replaces the selection.
"make it friendlier and change Tuesday to Thursday"Distill
Talk something through for five minutes. Review a short, ordered version before it goes anywhere.
Five minutes of thinking → a brief you'd sendTasks
Say what needs doing. Dates, tags and people stay with the task they were said with. Review, then copies go to Reminders, Obsidian, Microsoft To Do, ClickUp or Notion.
"call John about the invoice Friday, and ping design for the mockups"Meetings
Record a call or a room. Speakers are told apart, each turn transcribed, and you get a Markdown note, with a summary when cleanup is available. On the Mac it hears both sides of the call.
Mac: ~/Documents/VoiceMerge MeetingsiPhone: Files → VoiceMerge → Meetings
And it knows where you are
On the Mac, modes are keyed on the app, and in a browser on the site too: formal in mail.google.com, technical in linear.app, your own prompt anywhere you like. A personal dictionary helps the cleanup pass spell your names and terms.
Private by default. Cloud by choice.
Speech-to-text runs on the device by default. For the cleanup pass you pick a path, and every dictation shows which one ran.
On your device
Speech becomes text on the device: Parakeet or Whisper on the Mac; Whisper, Parakeet or Apple's recognizer on the iPhone. Cleanup runs on a small local model: Ollama on the Mac (free, a separate install), or Apple Intelligence on the iPhone. Apple's recognizer needs iOS 26 on an iPhone that supports it; Apple Intelligence needs iOS 26 on an iPhone 15 Pro or later. Works on a plane.
- Nothing leaves the device
- No account
- Good for most dictation
Your own API keys
Already have an Anthropic, OpenAI or Groq account? Paste a key. It's stored in the Keychain and the app talks to them directly. On the Mac you can also hand speech-to-text to Deepgram, Groq or OpenAI with your key; that's off unless you turn it on, and then the audio goes to them.
- Their bill, their terms
- No VoiceMerge account needed
- Screen context never goes remote
Claude through VoiceMerge
In the iPhone app, sign in with Apple, buy a credit pack, and cleanup, editing, Distill and meeting summaries run on Claude. Each request costs 4× the model provider's list price, and the ledger shows every charge. Not on the Mac: there it's Ollama or your own key.
- Standard (Haiku 4.5) by default; Best (Sonnet 5, ~2× cost) in Settings
- Only the text of a request is sent
- Not used for training
The details, with what each provider receives: privacy policy and security.
A Mac app and an iPhone app
Mac
A menu bar app. Hold a key anywhere — any app, any text field — and the text is typed at the cursor when you let go.
- Hold Right ⌥ to dictate, Right ⌃ to edit, Left ⌃ for tasks, and fn for Distill once you turn it on in Settings. All changeable.
- Quick-tap to lock hands-free; Esc cancels.
- Live preview while you speak; optionally type words as you go.
- Modes per app and per site, starting with a set you can change. Personal dictionary and text replacements.
- Meetings capture the mic and what the Mac plays (macOS 14.2+).
- Searchable history of your last 1,000 dictations, on your disk. Delete it whenever you like.
iPhone
Tap, speak, stop. The clean text is copied — or inserted into any app through the VoiceMerge keyboard.
- The VoiceMerge keyboard: a full keyboard with a Dictate button. It opens VoiceMerge to listen and inserts the text when you come back.
- Start from the Action Button, a Shortcut, or Siri.
- Dictate, Edit (from the clipboard), Distill, Tasks, Meetings.
- Apple's streaming recognizer on iOS 26 where the iPhone supports it; Whisper or Parakeet from iOS 18.
- Pick the tone on the keyboard: standard, casual, formal, technical or verbatim.
- Its own dictionary and task list, kept on the phone. The Mac app keeps its own; tasks sync between them only if you sign in and turn that on.
In TestFlight; App Store review is next. The keyboard types without Full Access; only dictation needs it, to receive the text from the app. The keyboard has no network code.
Free to use. Pay only for the cloud you choose.
VoiceMerge
Everything on the device. No account, no subscription, no trial that ends.
- Dictate, Edit, Distill, Tasks, Meetings
- Local speech and local cleanup
- Bring your own API keys if you want
VoiceMerge credits
Claude through VoiceMerge, in the iPhone app, at 4× the model provider's list price. Cleaning up a dictation costs under half a cent; summarizing an hour-long meeting, 5 to 10 cents.
- Buy in the iPhone app through the App Store
- A ledger of every request and what it cost
- Have a code? Redeem it in the app for free credit
- Web checkout by card, with optional auto-refill, isn't open yet
- Credits are on iPhone today; the Mac uses Ollama or your own key
Things people ask
Does anything leave my device?
Not unless you choose it. Speech-to-text and cleanup run on the device by default. If you add your own API key, or buy credits on iPhone, the text of a request goes to that provider; if you turn on task sync, your tasks and their notes are kept in your account. The screen context never leaves. Audio leaves only if you turn on a remote speech provider on the Mac with your own key.
How good is the on-device cleanup?
Good for ordinary dictation: fillers, punctuation, sentence fixes, matching tone. It's a small model, so long Distill and meeting summaries are where Claude is noticeably better. The app tells you which path ran on every dictation so you can judge for yourself.
Why credits instead of a subscription?
Most people never need the cloud tier, and those who do use it unevenly. A $10 pack lasts weeks for someone who dictates all day and summarizes a few meetings a week, and months for light use. The ledger shows exactly where it went.
What happens to my old dictations?
The last 1,000 stay in a history database on your device. The list shows the latest 500, and search covers all of them. Re-clean an entry, copy the raw words, delete one or all. Turn history off and nothing is kept.
Which languages?
Parakeet handles English and 24 other European languages. Whisper covers around 99. Cleanup follows the language you spoke.
Is it open source?
Not at the moment. Mac releases come from this site, signed with our Developer ID and notarized by Apple, and everything it does with your speech happens on your device unless you turn on a cloud provider. What we run is the service behind accounts: credits on iPhone, so you don't have to manage keys, and task sync, if you turn it on.