MacWhisper 15.8 — macOS

You have a two-hour interview recording on your desk and a client who wants a written summary by the end of the day. You open MacWhisper, drag the MP4 file onto the window, and pick a language. The transcription runs on your Mac, so the audio never leaves the machine. When it finishes, you export the text as a Word document and send it off.

What MacWhisper handles well

• Transcribes audio files in 100 languages, with automatic detection or manual selection.
• Supports drag-and-drop for MP3, WAV, M4A, MP4, MOV, OGG, and OPUS files.
• Exports subtitles as SRT or VTT, and documents as Word, PDF, or HTML.
• Records audio directly in the app for quick meetings or notes.
• Real-time meeting transcripts keep up as people talk.
• Folders and subfolders keep transcripts organized.

Transcription runs entirely offline

• Transcription is local, so no internet connection is needed after you download the app.
• Version 15.2.1 hides empty Local Models categories on Macs that can't run them, and dropped Qwen3-ASR support on M1-series Macs.
• Requires macOS 14 or later, works on both Apple Silicon and Intel 64-bit Macs.

Picking a model that's too large slows you down

The first one is easy to do: pick the wrong model size and then wonder why transcription takes forever. Beginners often grab Large-V2 for a short voice memo, when Tiny or Small would finish in a fraction of the time and give the same usable text for that kind of recording. The app lets you choose models for a reason — Tiny and Base are fine for straightforward speech with little background noise.

Another common slip is leaving language detection on for a clearly English recording. Automatic detection adds a few seconds of processing and occasionally picks wrong if the audio is short. If you know the language, set it yourself.

People also forget that speaker separation works best when you turn it on before recording, not after. If you start a meeting transcript and expect it to label speakers automatically, you need to check the setting first. The app does not guess after the fact.

And then there is the export step. Beginners sometimes export to Word and then spend time cleaning up line breaks that come from the transcript segments. MacWhisper lets you edit and delete segments before you export. A quick pass through the text in reading mode saves that cleanup work.

Start with a short file and the Small model

Start with a short file — a minute or two — to get a feel for the workflow. Pick the Small model, let the transcription finish, and try each export format. You will learn the rhythm of the app without waiting on a long recording.

Set up a folder for transcripts and another for meeting transcripts. The version 15.2 added folders and subfolders, so you can keep work projects separate from personal notes from the start.

Try the system audio transcription feature early. If you are in a meeting and cannot record from the microphone, this lets the app capture what comes through the speakers. Test it once when you have a few minutes to confirm it works with your setup.

Get comfortable with batch transcription. Drop three or four interview recordings into the app at once, select the same language and model, and let it process them in sequence. You can check the first one while the others run.

Copy only the quote you need from a transcript

You can copy all text at once or select only the sections you need. That sounds obvious, but it means you can pull one quote from a long transcript without exporting the whole thing.

The reading mode is useful for proofreading. It highlights each segment as you go through the text, so you can spot where the transcription might have misheard a word and fix it inline.

Editing and deleting transcript segments works after transcription finishes. If the app split a sentence awkwardly, you can merge or remove that segment before saving.

When you export subtitles, you get both SRT and VTT formats. That works for video editors who need one format and podcasters who need the other.

Built for journalists and podcast editors

Journalists and podcast editors are the obvious fit. A reporter who records interviews on a voice recorder can drag the files in, let the transcription run while they write notes, and export to Word or HTML for editing. Podcast editors benefit from the speaker separation and batch export — process a whole episode, separate the hosts, and pull out quotes for show notes.

Meeting-heavy workers also get value. Real-time meeting transcripts let you focus on the conversation instead of typing notes. The app runs locally, so nothing leaves your Mac, which matters if you discuss confidential material under an NDA or sensitive project details.

Lecturers and students who record lectures can transcribe them later in the language they need. The 100-language support covers everything from English to less common languages, and the local processing means no data goes to a cloud service.

Anyone who needs subtitles for video content will find the SRT and VTT exports useful. YouTube creators, corporate video teams, and training content producers can generate subtitles without sending audio to an online service.

You download it, you drag files onto it, you copy the text, and you move on.

• Improved speaker recognition with a new diarization model.
• Added real-time meeting transcripts.
• Added folders and subfolders for organizing transcripts and meetings.

• macOS 14 or later
• Supports Apple Silicon and Intel 64-bit Macs
• No activation required