FAQ

No, and it cannot. Detection only puts a cue on screen when it sees an app open the microphone. Every recording begins with a press, and if you ignore the cue it goes away and nothing is written.

It watches which application holds the microphone, so it works with about thirty services out of the box — video meetings, team chat and messaging apps — including calls in a browser tab. Anything missing you can add yourself, by application or by window title.

Yes. Your microphone and whatever your Mac is playing are mixed as it records, into one file. Headphones change nothing: the other side is taken from the audio itself rather than from a speaker. You can also record just your own voice, or just theirs.

The whole display, a single window, or an area you drag — with the audio or without it. A screen recording can have its audio pulled out afterwards into a file about a twentieth of the size.

~/Movies/RecordCue by default, in a folder per service, named with the date and the call. You can change the folder, the folder structure and the file name pattern in Settings. They are ordinary .m4a and .mp4 files — nothing proprietary, nothing to export.

Yes, using the speech recognition built into macOS. Turn it on for every recording, or transcribe a finished one later. Each passage is marked with which side was speaking — your microphone, the other end, or both at once. The result is a plain text file with timestamps, saved next to the recording.

No. Recognition is forced to run on the device. No audio is uploaded, to Apple or to anyone else.

It follows the language your Mac is set to, and uses whichever recognisers macOS has installed on the device. A call in a language your Mac is not set to will come out poorly — that is a limit of the recogniser, not of the recording, and the audio is untouched either way.

No. There is no account and no server. RecordCue makes no network requests of its own — no analytics, no telemetry, no crash reporting. The only connection involved in using it is the one the App Store makes to sell it.

Because macOS will not hand one app the sound another app is playing by any other route: the screen capture API is the only interface that exposes it. Recording your own microphone alone works without that permission. Nothing on your screen is captured unless you choose Whole screen, A window or An area.

Not in the sense people usually mean. It does not summarise, it does not extract action items, and it has no model of its own and no key to anyone else's. Transcription is the recogniser already in macOS, running on the device.

Everything, with one limit. Detection, the cue, every recording mode, transcription and filing are all in the free version and stay there. The first fourteen days have no limit at all; after that each recording stops at ten minutes and is saved intact.

It removes the ten-minute limit, and nothing else. It is a single purchase inside the app, not a subscription — bought once, no renewal, no account.

No. The purchase belongs to your Apple Account, so every Mac you are signed in to has it, and it comes back after a reinstall.

macOS 13 or later, on a Mac with Apple silicon.

It depends on where everyone in the call is. In many places you have to tell the other people; in some you need their agreement first. RecordCue cannot know where anyone is and does not decide this for you. Ask first — it is also the thing that keeps the recording worth having.

Yes. A read-only command ships inside the app: it prints what RecordCue knows about a finished recording as JSON — where the file is, when it was made, which call it was and which side was speaking — so anything that can run a shell can find it without you hunting for a path. Settings → AI tools provides a setup request for your local AI agent, and the Agent setup page walks through it. For things that should happen on their own after a call — a Slack summary, a note in Obsidian, action items in Reminders — see the automation recipes.