Convert audio to text on a Mac

inlaut transcribes audio and video files on your Mac: interviews, voice memos, lectures or the recording of a meeting. No upload, no account and no minute packs.

As of 20 October 2026, inlaut 0.5.1. This page is written by the developer of inlaut, a free dictation app for the Mac.

How it works

1. Click the inlaut icon in the menu bar and choose “Transcribe File …”.
2. Pick the audio or video file.
3. The transcript appears in its own window. Turn on timestamps if you need them, copy the text or save it as a text file.

inlaut does not save the transcript on its own. Save it before you close the window.

Which files work

• Audio: MP3, M4A (for example from the Voice Memos app), WAV and other formats macOS can open.
• Video: MP4 and MOV. inlaut only reads the audio track.
• Browser recordings: WebM with Opus audio, which macOS itself cannot open.

Other codecs in WebM or Matroska (MKV) and Windows Media are not supported. inlaut then tells you which formats work.

How fast, how long

Recognition runs on your Mac’s Neural Engine. In one test inlaut took five to eight seconds for just under five minutes of audio. Long files are split at quiet moments into sections of 45 to 75 seconds and processed one after another. That keeps memory use small, and silent sections are skipped. If you dictate in between, the dictation waits for the current section at most.

What the transcript contains

• Continuous text in English, German or both mixed, one paragraph per section.
• Optionally a timestamp at the start of each section.
• Your replacements apply here too, so names and jargon come out right.
• Voice commands and snippets do not apply: if someone in the interview says “new line”, it appears as text.
• No speaker labels, no summary.

Who it is for

• Journalism and research: transcribe interviews without handing the recording to a service.
• Students: make lectures and your own voice notes searchable.
• Podcasts and video: a text version as a starting point for show notes or subtitles.
• Office: a meeting recording as the basis for the minutes.

Only record conversations when everyone involved agrees.

Limits

• English and German only.
• Speakers are not identified. Overlapping voices or a noisy recording make recognition worse.
• No live meeting transcription: inlaut transcribes finished files.
• Mac version only, for Apple Silicon with macOS 26 or later. The Windows beta does not transcribe files.

Get to know inlaut

More ways to use inlaut

• Dictate emailsGreeting, paragraphs and sign-off as you speak.
• Dictate promptsSpeak long prompts for ChatGPT, Claude and Cursor.
• Dictation apps comparedTen Mac apps side by side.

Frequently asked questions

Is my recording uploaded?

No. inlaut transcribes the file on your Mac. There is no upload, no account and no minute quota.

Which file formats can inlaut transcribe?

Audio and video files macOS can open, such as MP3, M4A, WAV, MP4 and MOV, plus WebM with Opus audio. Other codecs in WebM or Matroska and Windows Media files do not work.

Does inlaut tell speakers apart?

No. The transcript is continuous text without speaker labels. Optionally, each section starts with a timestamp.

How long does transcription take?

In one test inlaut took five to eight seconds for just under five minutes of audio. Long files are split into sections of about a minute and processed one after another.

Notes

Product and company names are trademarks of their owners and are used only to identify the products. inlaut is not affiliated with any of these companies.

Something out of date or missing? Write to hallo@inlaut.de.