Offline Transcription Software: Apps That Keep Audio on Your Computer
Data checked:
Cloud services upload your audio to transcribe it. The apps below run the speech-recognition model on your own computer instead, so recordings are not uploaded for transcription and there is no monthly minute allowance. The trade-off is speed: how fast they run depends on your hardware and the model you pick.
| Tool | Platforms | Paid from | Speaker labels | Export | Languages |
|---|---|---|---|---|---|
| Aiko File transcription, Offline / on-device | Mac, iOS | $24 one-time | No | SRT, CSV, JSON | ENJAZH 100 languages |
| aTrain File transcription, Offline / on-device | Windows, Linux, Mac | Not verified | Yes | Not verified | ENJAZH 99 languages |
| AudioPen File transcription, Live transcription, Offline / on-device | Web, iOS, Android, Mac, Windows, Chrome | $8.25/mo billed yearly (Prime) | No | Not verified | ENJAZH |
| Buzz File transcription, Offline / on-device | Mac, Windows, Linux | Free | Yes | TXT, SRT, VTT | ENJAZH 100 languages |
| MacWhisper File transcription, Offline / on-device | Mac | €64 one-time (Pro) | Yes | TXT, SRT, VTT, DOCX, PDF, MD, HTML, CSV | ENJAZH 100 languages |
| noScribe File transcription, Offline / on-device | Windows, Mac, Linux | Not verified | Yes | HTML, TXT, VTT | ENJAZH |
| superwhisper Live transcription, File transcription, Meeting notes, Offline / on-device | Mac, Windows, iOS, Android | $8.49/mo $7.08/mo billed yearly $249.99 one-time (Pro) | Paid plans | Not verified | ENJAZH 100+ languages |
| Texter File transcription, Live transcription, Offline / on-device | Mac, iOS, Android | ¥1,800/mo ¥733/mo billed yearly (Premium) | Paid plans | TXT, PDF, SRT, MD, HTML | ENJAZH 99 languages |
| Vibe File transcription, Offline / on-device | Mac, Windows, Linux | Free | Yes | SRT, VTT, TXT, HTML, PDF, JSON, DOCX | ENJAZH |
| whisper.cpp File transcription, Offline / on-device, API | Mac, Windows, Linux, iOS, Android, Web | Not verified | Not verified | Not verified | ENJAZH |
On a Mac
- MacWhisper has a free version and a one-time Pro licence (€64). It lists automatic speaker recognition, records Zoom, Teams and Webex meetings from your Mac without a bot, and exports SRT, VTT, DOCX, PDF, CSV and more. It needs macOS 14 or later. If you switch it to a cloud model, your audio goes to that provider.
- Aiko is a one-time $24 purchase for Mac and iPhone with a 14-day TestFlight trial. It is simpler and does not label speakers.
On Windows or Linux
- Vibe is free and open source (MIT). It does batch transcription and speaker diarization and exports SRT, VTT, TXT, HTML, PDF, JSON and DOCX.
- Buzz is free and open source (MIT) as well, with speaker identification and TXT, SRT and VTT exports. Current Mac builds need Apple Silicon.
Languages
All four use OpenAI’s Whisper models, whose language list has 100 entries including Japanese, Chinese and Cantonese. Vibe also offers Parakeet and Nemotron models. Accuracy varies by language and model size; we have not tested it yet.
When a cloud service is the better choice
On-device apps generally don’t offer meeting bots or shared team workspaces, and long files take as long as your hardware needs. If you want a hosted service that still keeps your files out of AI training, compare Sonix, tl;dv and Fireflies.ai in the directory.