MacWhisper
Mac transcription app that turns audio, video and meetings into text locally, without uploading your recordings to a cloud.
What it does
MacWhisper turns audio and video files into text right on your Mac. You drag files into the window, record from the microphone or paste a YouTube link. It recognises around 100 languages and separates speakers automatically. You can export subtitles (SRT, VTT), plain text, Markdown, PDF or Word. It also offers system-wide dictation, recording of Zoom, Teams or Webex calls without a bot joining, watched folders, a command-line interface, and hand-off to Notion, Obsidian, Zapier, Make or n8n.
Who it suits
Anyone who regularly needs interviews, meetings, podcasts or videos in writing and doesn’t want to upload confidential recordings, for example in consulting, medical practices, newsrooms or research.
Data protection for businesses
By default, transcription runs on the Mac with local models such as Whisper or Parakeet, and recordings stay on the device. Data only leaves the device if you turn on cloud transcription (e.g. Deepgram, ElevenLabs), DeepL translation or AI providers such as OpenAI. Their terms then apply, including on training. Summaries can also run locally through Ollama or LM Studio. The vendor does not offer EU data hosting and does not mention a data processing agreement. For teams there are licence packs of 5 to 50 seats, which you can roll out via MDM.
Limits
The directly purchased version runs on macOS only. The App Store variant “Whisper Transcription” for Mac and iOS does less and is sold as a subscription. Speaker recognition, export to formats other than text, batch processing, meeting recording and your own AI providers are Pro only. The large models need more than 8 GB of memory. The privacy policy dates from 2024 and says nothing about the cloud features; the vendor gives no company address.