Who said what
Recordings with both tracks get speaker labels from the audio itself: your side is "Me", the other side is labelled separately. Works with every transcription provider, entirely on your Mac.
Stenobar
Your meetings, recorded on your Mac.
No account, no server, no subscription. Their audio and yours on separate tracks, so the transcript always knows who said what, and the recording stays a file you own.
Native Swift & SwiftUI Requires macOS 26 or later on Apple Silicon v1.0.0-beta.29
What's new in each release → Catch what you hear. Keep what you say.
Stenobar is a recorder first. Transcription and summaries are optional: they run on-device, or against your own cloud keys, or you never turn them on.
Record
Capture system audio and your mic as two separate tracks, straight from a menu-bar item. The Start button says exactly what it will capture, Sound and Microphone are two cards with their own switches, and Test levels checks a source is picking up audio before you commit. Grab everything the system plays, or just the apps you pick: Zoom, Safari, whatever you choose.
Never miss a call
When Zoom, Teams, Meet, or another meeting app takes the mic, Stenobar asks whether to record, names the app, and pre-selects it. Yes, No, or Ignore for now, with a project picker right on the card. It waits for an answer instead of disappearing. Stenobar only checks which apps are on the mic; it never listens in.
Transcribe Optional
Turn any recording into searchable text. Run it entirely on-device with Apple Speech, WhisperKit, or Parakeet, or point it at a cloud engine: AssemblyAI, Deepgram, ElevenLabs, Soniox, OpenAI, Groq, Google Cloud, or Gemini. Completely optional: with nothing configured, your recordings are still yours to play, export, and keep.
Summarize & tag Optional
Hand a transcript to an LLM for a summary and suggested tags, whether that's Apple Intelligence on-device, a local Ollama model, or Anthropic, OpenAI, Gemini, OpenRouter, or any OpenAI-compatible endpoint. It's optional too, and separate from transcription, so you can mix providers or skip it altogether.
Dictate
Hold a hotkey, or just a single modifier key, and speak. A floating HUD shows live text, transcribed on-device or streamed through the cloud engine of your choice, then pasted straight into whatever app you are using. Every take is kept in a searchable Dictation History.
Thoughts
A second hotkey for fleeting ideas: speak, and an LLM files the result where it belongs. Apple Reminders, Apple Notes, Todoist, TickTick, Things, Obsidian, Bear, Notion, a Shortcut, or a local inbox.
Search everything
A three-pane library with full-text search across titles, tags, summaries, and transcript bodies. Filter by date, sort, star, and group by project.
The windows behind the menu bar.
Recording lives in the menu bar. Everything it produces lands in a few plain macOS windows.
The details that keep it solid.
Crash-safe
Every session is written as raw PCM WAV with the header flushed every few seconds, so a crash still leaves usable audio behind.
Projects & presets
Group recordings into project folders, each with its own transcription and summary overrides. Save a source, mic, and project as a one-click preset.
Your format
16- or 24-bit, 44.1 or 48 kHz, mono or stereo. Export audio to M4A or FLAC and transcripts to TXT, Markdown, SRT, or VTT.
Combined track
Optionally stitch system and mic into one mixdown (mono, split stereo, or mixed), and re-stitch any time you change your mind.
Import audio
Drop in existing WAV, MP3, M4A, AIFF, FLAC, or CAF files to transcribe, tag, summarize, and search alongside your recordings.
Auto-stop & retention
Optional auto-stop after a stretch of silence, with a countdown you can cancel. Optional retention trashes old, unstarred recordings after a set number of days.
Automate it
Start, stop, mark, dictate, and capture thoughts from Shortcuts, Siri, or Spotlight, or from any tool that can open a stenobar:// link.
How is this different from what you already use?
Fairly, we hope. Every detail comes from each vendor's own site, and every app gets a row for what it does better.
| Feature | Stenobar | MacWhisper | Granola | superwhisper | Otter |
|---|---|---|---|---|---|
| Built for | Recording calls you keep as audio | Transcribing files, plus dictation | Meeting notes for teams | Dictation | Team meeting notes, with a bot |
| Records a live call from your Mac | Yes | Pro only | Yes | Yes | A bot joins the call |
| Their audio and yours as separate tracks | Always. Who said what comes from the recording itself | Not stated | One mixed stream | Not stated | Not stated |
| Where transcription runs | On your Mac, or a cloud engine with your own key | On your Mac, or cloud providers with your own key on Pro | Granola's cloud | On your Mac, or cloud models on Pro | Otter's cloud |
| Account required | No | No | Yes | Not stated | Yes |
| Keeps the audio | Yes, as WAV files on your disk | Yes | Keeps notes, not audio | Not stated | In Otter's cloud |
| Minute or retention limits | None | None | Notes older than 30 days need a paid plan | None stated | 300 minutes a month and 30 per conversation on the free plan |
| Bring your own API key | Yes, at no extra charge | Pro only | No | Pro only | No |
| Dictation into any app | Yes | Pro only | No | Yes, it is the whole product | No |
| Price | Free while in beta | Free tier. Pro is €64 once | Free tier. Business is $14 per user a month | Free tier. Pro is $8.49 a month, with yearly and lifetime options | Free tier. Pro is $8.33 per user a month billed yearly, $16.99 monthly |
| Where it wins | Two tracks, no account, and the recording is yours. Needs macOS 26 on Apple Silicon | Years of polish, batch file transcription, and a wider choice of models | Polished notes with calendar context, team sharing, and Windows, iOS, and Android apps | The most refined dictation experience, on Mac, Windows, and iOS | Joins meetings you cannot attend, with web and mobile apps and team workspaces |
Spotted a mistake? Open an issue and we will fix it.
Bring your own keys, at no extra charge.
Stenobar doesn't tie you to one vendor. Run everything on-device, or drop in your own keys for the cloud services you already use, per recording, and only when you choose. Transcription and the language model are picked separately, so you can mix and match freely.
On-device
On-device transcription
- Apple Speech
- WhisperKit
- Parakeet
Whisper tiny.en → large-v3 · Parakeet TDT 0.6B (v2 / v3) with on-device speaker labels
Cloud
Cloud transcription
- AssemblyAI
- Deepgram
- OpenAI
- Groq
- Google Cloud
- ElevenLabs (beta)
- Soniox (beta)
- Gemini (beta)
AssemblyAI, Deepgram, ElevenLabs, Soniox, and Gemini also stream live dictation
On-device
On-device AI
- Apple Intelligence
- Ollama
Any local Ollama model, pulled from inside Settings
Cloud
Cloud AI: summaries, tags & thoughts
- Anthropic
- OpenAI
- Gemini
- OpenRouter
- OpenAI-compatible
OpenRouter includes a searchable catalogue with pricing, Auto Router, and spending limits
Thoughts
Thoughts destinations
- Apple Reminders
- Apple Notes
- Todoist
- TickTick
- Things
- Obsidian
- Bear
- Notion
- Shortcuts
- Stenobar Inbox
Lists are rendered natively: Todoist sub-tasks, TickTick and Things checklists, Notion bullets
Your recordings stay on your Mac.
Audio never leaves your Mac unless you supply your own API key. On-device transcription, speaker labels, and Apple Intelligence run entirely on the device. Cloud providers only ever touch the single recording you ask them to process, and the app tells you what each one does with your data before you save a key. Your API keys live in the macOS Keychain, nowhere else. No analytics. No tracking. No account.
- On-device by default
- You keep the audio
- Opt-in cloud, per recording
- Zero telemetry
Coming with 1.0
Stenobar Pro.
A one-time purchase, no subscription, for the automation around the recording. The recorder itself stays in the free tier, complete, for as long as you like. While Stenobar is in beta, everything is on for everyone.
Free tier, forever
- Records system audio and your microphone as separate tracks
- Per-app capture: only Zoom, only the browser, or everything
- On-device transcription with Apple Speech, WhisperKit, or Parakeet
- Who said what, from the two tracks, on-device
- On-device summaries with Apple Intelligence
- Bring your own API keys for any cloud provider, at no extra charge
- Every export format: WAV, M4A, FLAC, TXT, Markdown, SRT, VTT
- The full library: search, projects, tags, markers, playback, retention
- Hold-to-dictate into any app
- Thoughts capture, routed to Reminders, Notes, Shortcuts, or the Inbox
Pro adds
- Records your meetings automatically, without you opening anything
- Names the meeting from your calendar, and reminds you before it starts
- Your prompt library, per-project summary settings, and transcript tuning
- Thoughts to Todoist, TickTick, Things, Obsidian, Bear, and Notion
- Two Macs, and every 1.x update
Pro is on for your first 21 days. When it ends, the recorder keeps working, your library stays open, and every export format still works. You lose the automatic meeting capture, not your recordings.
- No minute limits, no retention limits, no recording counts. Ever.
- The price and the final Pro line-up will be published here before 1.0 ships.
Catch the next thing you'd have lost.
Free while in beta.
What's new in each release →