A Mac audio transcription application with free local text and subtitle output and a paid Pro edition.
Freemium
noScribe
A free local interview transcription application with speaker labels, pause notation and an audio-connected editor.
About noScribe
noScribe is a free, open-source desktop transcription tool for interviews and other sensitive audio data, and is unrelated to the paid transcription service sold at the noscribe.ai domain. It runs entirely locally on the user’s computer with no cloud processing, using OpenAI’s Whisper and Hervé Bredin’s Pyannote models for transcription and speaker recognition, with support for roughly 60 languages. It includes an audio-linked editor for transcript correction, and outputs HTML (compatible with MS Word), plain text, or WebVTT files. Specialized features include pause and overlapping speech notation, optional filler word transcription, and timestamp insertion. Local processing means one hour of audio takes 1–3 hours to transcribe (longer on older machines). The ~3GB installer includes the required AI models, and an NVIDIA CUDA-enabled GPU can speed up processing. Post-transcription review and correction are always required.
Key features
- Local, no-cloud processing of audio and video files
- Speaker recognition and labeling for transcripts
- Pause and overlapping speech notation support
- Audio-linked editor for transcript review and correction
- Output to HTML, plain text, or WebVTT format
- Supports approximately 60 languages and dialects, available for Windows, macOS, and Linux
Best for: It is best suited for users handling sensitive audio data, including qualitative social researchers, journalists, medical professionals, and NGOs, who require private, local transcription without cloud uploads.
Good for
- Free, open-source (GPL-3.0) with no cloud processing for user privacy
- High-quality transcription powered by Whisper and Pyannote AI models
- Includes specialized qualitative research features like pause notation and optional filler word transcription
Watch out for
- Local processing means 1 hour of audio takes 1–3 hours to transcribe, longer on older machines
- Installer size is approximately 3GB as it includes required AI models
- Post-transcription review and correction are always necessary
noScribe at a glance
- Website
- noscribe.de/en
- Category
- AI & Machine Learning › AI Speech Transcription
- Type
- Online tool
- Pricing
- Free
- Open source
- Yes
- Last checked
- Oct 9, 2026 · Online when checked
- Listed
- Oct 9, 2026
noScribe alternatives
Tools and sites you could use instead, ranked by how closely they match.
Popular in AI Speech Transcription
Well-known listings from the same topic.
Transgatetransgate.aiWavewave.coRev.airev.aiGood Tapegoodtape.ioTalkNotestalknotes.ioSpeech to Notespeechtonote.com
noScribe FAQ
Is noScribe free to use?
Yes, noScribe is free and open-source, and will always remain free.
Does noScribe upload audio data to external servers?
No, all transcription processing happens locally on the user's computer with no cloud involvement.
What file formats can noScribe export?
It can export HTML files compatible with MS Word, plain text, or WebVTT subtitle files.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.