How do you transcribe an audio recording you already have?

Cloud service or on-device app

There are two ways to turn a recording into text, and the choice mostly comes down to how sensitive the audio is. A cloud service is quick to try: you upload the file, it processes on their servers, and you download the text. The trade-off is that a copy of your recording now sits with a third party, sometimes retained until you delete it.

An on-device app does the same job on your own Mac, so the recording stays yours. For anything private, an interview under embargo, a legal call, a personal note, that is the safer default.

Getting the file ready

Most recordings are already in a format a transcription tool can read. Audio from a phone is usually m4a, a voice recorder often gives you wav or mp3, and a screen capture is typically mp4 or mov. A good on-device tool will pull the audio track out of a video for you, so you do not need to convert anything first.

If your file is in an unusual format, converting it to wav or mp3 with a free tool is a safe fallback, but it is rarely necessary.

Running the transcription

The steps are the same whatever tool you pick. Open the app, import or drop in the file, and let it process. On-device transcription runs on your Mac’s own chip, so the time depends on the length of the recording and your machine, but a normal meeting is usually done in a few minutes and runs in the background.

When it finishes you get the text, ideally with timestamps so you can jump back to any moment in the audio to check it.

From transcript to something usable

A raw transcript is long. If you want notes rather than a wall of text, the next step is a summary. Some on-device tools can write one for you on the same machine, pulling out the key points and any action items, without sending the transcript anywhere. Otherwise you can paste the text into a model of your choice, keeping in mind that a cloud model means the text leaves your machine.