Drop your audio. Transcript in seconds. 30 free min, then $2 = 200 min
MacWhisper
The polished native-Mac app for local Whisper transcription — drag-and-drop file editor, speaker diarization, system-wide dictation, AI summaries, and SRT/VTT/PDF export. macOS only, by Jordi Bruin.
MacWhisper is a native macOS app that wraps OpenAI's Whisper models behind a Mac-style editor. Transcription runs locally on Apple Silicon (or Intel) with Metal acceleration — your audio stays on the machine. Optional cloud AI hooks (OpenAI, Anthropic, ElevenLabs, Groq) handle summaries, rewrites, and translations only when you opt in.
Best for journalists, podcasters, researchers, and Mac power-users who want a polished editor and system-wide dictation without a terminal. The Free tier covers smaller Whisper models; Pro ships as a one-time Gumroad license (large-v3, diarization, batch, dictation, AI services); Whisper Transcription on the App Store offers a monthly/annual subscription path. Check the Gumroad listing for current numbers.
What it is
MacWhisper is the most polished Mac-native Whisper front-end. Free tier covers small models; the Pro license unlocks large-v3, diarization, and batch. Perfect for journalists, researchers, and indie podcasters who want a file-drop tool that just works on macOS.
Watch out for: Mac-only; closed-source; diarization + large models are paid tier.
Install / use
Download MacWhisper goodsnooze.gumroad.comWhere MacWhisper fits
Six use cases the app is purpose-built for. Every card links to the corresponding feature page on macwhisper.com or the canonical docs.
Drop an interview, get a clean transcript with click-to-seek playback. Edit the text inline, search across long sessions, and export the result with timestamps. Works on multi-hour files locally — no upload, no queue, no per-minute fee.
Local model · Metal acceleration on Apple Silicon
Record system audio or import a Zoom/Teams export, run diarization to label Speaker 1 / Speaker 2 / Speaker 3, then trigger an AI summary or action-item extraction via the built-in service hooks. Useful for 1:1s, panels, and customer calls.
Exports group automatically by speaker
Generate SRT and VTT subtitle files from any audio or video. The inline video player can display multiple translated subtitle tracks side by side — handy for proofreading captions across languages before publishing to YouTube or your CMS.
Bulk export across batch jobs
Hold a global hotkey in any text field — Slack, Mail, VS Code, Notion — and MacWhisper transcribes your voice with Whisper-grade accuracy, then pastes the result inline. Optional AI prompting can clean up filler words or rephrase before paste.
Pro feature · runs entirely locally by default
Post-process any transcript with bring-your-own-key calls to OpenAI, Anthropic, Groq, or ElevenLabs. Summarize, extract bullets, translate, clean dictation, or run a custom prompt template. Cloud calls only fire when you trigger them — transcription itself stays local.
Pro feature · prompts are user-editable
Queue a folder of mixed-language recordings, let MacWhisper auto-detect each file's language, and walk away. The queue view tracks progress; results export per-file or as a combined archive. Pairs well with a Hazel rule for fully-automated ingestion.
Pro · large-v3 yields the best non-English accuracy
Getting started in 3 steps
MacWhisper is a Mac app — onboarding is a download and a few permission grants, not a terminal session. Walk through these three steps in order.
Grab the free build from macwhisper.com, drag an audio or video file onto the window, pick a model, and hit transcribe.
1. Visit https://macwhisper.com and click Download.
2. Open the .dmg, drag MacWhisper.app to /Applications.
3. Launch MacWhisper.
4. On first run, pick a Whisper model:
· Tiny / Base → fast, free, fine for short clips
· Small / Medium → balanced (free tier)
· Large-v3 → best accuracy (Pro)
5. Drag an audio or video file onto the window.
6. Hit Transcribe — the file processes locally on your Mac.
7. When done: edit inline, then File → Export → pick SRT / TXT / VTT / PDF.Set up the global hotkey and grant Accessibility so MacWhisper can paste into any text field — Slack, Mail, VS Code, Notion, anywhere a cursor blinks.
1. MacWhisper → Settings → Dictation tab.
2. Toggle 'Enable Dictation'.
3. Pick a Whisper model for dictation
(Small is a good default — fast, accurate enough for chat).
4. Bind a global hotkey (default: hold Option twice, or set your own).
5. macOS will prompt for:
· Accessibility permission (paste into other apps)
· Microphone permission
Grant both in System Settings → Privacy & Security.
6. Click into any text field, hold the hotkey, speak.
7. Release — MacWhisper transcribes locally and pastes the result.
8. Optional: turn on 'AI cleanup' to strip 'um/uh' and tidy grammar via your
OpenAI / Anthropic / Groq key before paste.Queue a folder of files for unattended transcription. Set output format and destination once, then drop new files in over time and let the watch folder pick them up.
1. Open MacWhisper → File → Batch / Queue.
2. Drop a folder (or multiple files) into the queue view.
3. Pick the model + language (or leave language on Auto-detect).
4. Set the default export format:
· SRT → subtitles for video editors
· VTT → web video / Whipscribe-compatible
· DOCX → drop into Google Docs / Word
· CSV → speaker-labeled rows (with diarization on)
· PDF → distributable transcript
5. Pick the export folder (next to source, or a fixed location).
6. Start — MacWhisper processes the queue serially.
7. Optional: pair with a Hazel rule or Folder Action so any new file dropped
into ~/Inbox/Audio auto-launches MacWhisper and queues it.Features
| Speaker diarization | Yes |
| Word-level timestamps | Yes |
| Streaming / real-time | No |
| Languages supported | 99 |
| HIPAA eligible | No |
Links
- macwhisper.com ↗product homepage — features, screenshots, system requirements
- goodsnooze.gumroad.com/l/macwhisper ↗official Gumroad sales page — one-time Pro license by Jordi Bruin / Good Snooze
- MacWhisper release notes ↗version history — recent builds include CLI control, queue toolbar refresh, ElevenLabs fixes
- macwhisper.helpscoutdocs.com ↗official user manual — install, dictation, diarization, batch, AI services, export formats
- MacWhisper vs Whisper Transcription ↗explains the Gumroad app vs the App Store subscription variant from the same developer
- @jordibruin on X ↗developer account — release announcements, feature previews, discount codes
- @jordibruin@mastodon.online ↗Mastodon mirror — same release announcements, Fediverse-native
- goodsnooze.gumroad.com ↗Good Snooze storefront — other indie apps by the same developer
MacWhisper vs Whipscribe
| Feature | MacWhisper | Whipscribe |
|---|---|---|
| Category | Desktop apps | Transcription APIs |
| Pricing | freemium | free beta |
| Speaker diarization | Yes | Yes |
| Word timestamps | Yes | Yes |
| Streaming | No | No |
| Languages | 99 | 99 |
| Platforms | macOS | Web, API, MCP |
Alternatives to MacWhisper
Frequently asked about MacWhisper
Is MacWhisper free?
MacWhisper has a free tier that covers smaller Whisper models and basic transcription. Large-v3, speaker diarization, and batch processing live in the paid Pro tier. Check the Gumroad page for current pricing.
Does MacWhisper work on Windows?
No. MacWhisper is macOS-only. On Windows or Linux, the closest equivalents are Buzz (cross-platform, free, open-source) or any of the OpenAI Whisper API clients.
Does MacWhisper run locally or send data to the cloud?
Transcription runs locally on your Mac using Whisper — your audio does not leave the machine by default. The app also has optional integrations with cloud APIs if you configure them.
Does MacWhisper support speaker diarization?
Yes, on the Pro tier. Diarization labels 'Speaker 1 / Speaker 2' in the transcript. The free tier transcribes text-only.
Can MacWhisper transcribe YouTube URLs?
Recent versions support URL input (YouTube, etc.) in addition to file drop. Feature availability varies by version — check the MacWhisper changelog for current support.
Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.