Dictation and a note taker: one app or two?

Dictation and a note taker do different jobs. Dictation puts your own words into whatever field you are typing in; a note taker runs while people talk to each other and hands back who said what, with a summary. You can cover them with two apps or with one. Dalm is one Mac app that does both: the speech is recognised on your Mac for either job, no audio is sent anywhere, and the text is synced to our servers automatically. What differs between the two jobs is who sees the text afterwards and whether a language model reads it. There is a Free plan.

Two jobs: typing by voice and keeping a record of a conversation

One is you talking to your computer instead of typing, and nobody else is part of it. The other runs while people talk to each other, and the work is in telling the voices apart and deciding what mattered. What sets the two apart is explained on a page of its own; this page is about whether they belong in one app.

Situations where you need both jobs

• A call, then the reply. You record the call and stop it when it ends; then you dictate the follow-up into your email. The call is a meeting: its transcript is synced to our servers, and a language model makes the summary. The reply is a dictation: it is typed into the email and kept in your list of dictations, and no model reads it on our side.
• A conversation in the room. Nothing to join: the app records through the microphone and tells the voices apart on your Mac. It is a meeting like any other, and its text is synced to our servers the same way.
• A phone on speaker. It is recorded the same way as a conversation in the room — the phone is simply one more voice at the table.
• A request to a model. You dictate a prompt into the chat of whatever AI tool you use. It is a dictation: the words land in that chat field, and whatever you send goes to that tool. Your own copy stays in your dictations; Dalm does not pass it to a language model.
• A manager with a team. Team meetings are seen by the whole workspace by default, and a meeting can be set to elevated or protected privacy. Your dictations — notes to yourself about a hire, a draft of a hard email — are seen by nobody but you, and if you leave the team they leave with you while the meetings stay. How a team workspace is set up.
• A recording from somewhere else. A voice memo or a video call someone sent you: bring the file in and choose whether it is a dictation or a meeting, and it follows that job’s road: as a dictation, only you see it and no model reads it on our side; as a meeting, it gets a summary like any other. The file is read where it is, not copied.

Each job, step by step

Dictation

1. Hold the right ⌘ — or the key you chose — and speak. For something long, tap it once to start and again to stop.
2. Let go. The speech is recognised on your Mac, terms from your dictionary are put right, and the text is typed into the field where your cursor is, without touching the clipboard.
3. The audio is deleted once the text exists. The dictation is kept in your list and synced to our servers.

Meeting

1. Start it from the menu bar, with ⌃⌥M, or from the prompt that appears when a calling app takes the microphone. Nothing joins the call: the app records your microphone and the sound of other apps.
2. While it runs, write your own notes in the notes field; the summary is built around what you wrote.
3. When you stop, the speech is recognised and the voices are told apart on your Mac. The audio is deleted once the text exists.
4. The transcript is synced to our servers, and a third-party language model makes the summary from that text; each point in the summary keeps the words it came from.
5. Naming the voices is optional. Until you answer, a guess from the model or “Speaker 1” stands in the name’s place.

Three ways to cover both jobs

	Dictation built into macOS + a note taker	A dictation app + a note taker	Dalm: one app for both
Apps to install and sign in to	One: the note taker	Two	One
Dictionaries of names and terms	For dictation: not stated in Apple’s guide. The note taker: its own, if it has one	One in each app that has one	One, used by both jobs
Where to search what you said	The note taker’s search; dictated words are wherever you typed them	Two places, one per app	One field, ⌘K, for both
Where the audio goes	Dictation: your Keyboard settings show whether it is processed on your Mac, and you choose whether Apple may keep audio. The note taker: its own documentation	Each app’s own documentation — two to read	Recognised on your Mac; no audio is sent anywhere
What happens to the text	Dictation: into the field; keeping it is not stated in Apple’s guide. The note taker: its own documentation	Each app’s own documentation	Synced to our servers automatically, for both jobs
Who sees it, and which model reads it	Each part by its own rules	Each app by its own rules	Differs by job — see below

“Not stated” means Apple’s guide does not describe it, not that it does not exist. Two apps can be the right choice; the table only shows what each way leaves you with. Dalm is not the only app that does both: some note takers now include dictation in their desktop app, and where the audio of that dictation goes is set out in their own documentation.

When two separate apps are the better answer

• You dictate now and then, and never need the words again. The dictation built into macOS may be all you need. What a dictation app adds on top.
• You need to hear a meeting again. Dalm keeps no recording to play back, so a tool that leaves you a sound file is the right one: recorders, by what you are left with.
• The text has to stay on your computer too. Dalm syncs the text of meetings and dictations to our servers automatically, and there is no setting that turns this off — so it is not the tool.

Where the two jobs part ways: who sees the text, and which model reads it

Up to the text, both jobs take the same road: the same speech model on your Mac, one dictionary, one account, and the text of both synced to our servers. After that they part.

	Dictation	Meeting
Voices	Yours, and only yours	Several, told apart on your Mac. Until you name a voice, a guess from the model or “Speaker 1” stands in its place
The audio	Not sent anywhere; deleted once the text exists	Not sent anywhere; deleted once the text exists
The text	Synced to our servers automatically	Synced to our servers automatically, with the names you gave the voices and your notes
A language model	Never: dictation goes through recognition and your dictionary only	When a summary, a document or an answer is made, the text is sent to a third-party language model; the audio never is
In a team	Nobody sees it but you — not a moderator, not the owner	The whole team sees it by default; elevated or protected privacy narrows that
If you leave the team	Your dictations go with you	The meetings stay with the team

The voice prints that tell one voice from another stay on your Mac and are not synced; the names you give the voices are. In a team, the dictionary is shared by the workspace, so a term is spelled the same way for everyone.

What to know before you choose one app

• There is no audio to listen to. The sound is deleted once the text exists; there is no player and no audio export.
• The text is synced to our servers automatically, for dictations and meetings alike, and syncing cannot be switched off.
• An account is required. Without signing in there is no window and no dictation.
• Mac only, with Apple Silicon · macOS 15.0 or later.
• Questions go to one meeting at a time. Asking one question across everything you have recorded is in development.
• One job at a time. While a meeting is recording, dictation does not start — the two share the microphone — so stop the recording to dictate.
• Voices are identified afresh in every meeting. A colleague you named last week is a voice to name again this week, on the same Mac too.

One archive for everything you say

Everything you say lands in one archive you can search: meetings, phone calls, conversations in the room, and what you dictate to a model.

In the summary of a meeting, each point keeps the words it came from. Every quote shows you where it was said — open it and the transcript opens at that place. The audio is gone by then, so the place is in the text, not in a recording.

Free, Pro or Team: both jobs start on Free

Dictation and meetings are both part of the Free plan, and Free is a plan with no end date, not a trial. Dalm needs a Mac with Apple Silicon · macOS 15.0 or later and an account; while the app is not public, the way in is the waitlist. What the paid plans add, and what they cost, is on the pricing page. For a group, the split above is the point: meetings are shared with the workspace, dictations stay with whoever made them — what a team workspace shares.