For macOS

Speech to Text.
Private. Instant.

Transcribes your voice — completely offline, right on your Mac. No cloud. No data ever leaves your device.

Download

One-time purchase. No subscriptions. Yours forever.

How it works

From recording to text in seconds.

⌘ ⇧
1

Press the hotkey

Hit ⌘⇧Space from any app to start recording.

2

Speak

WhisperBar records and shows you a live waveform as feedback.

3

Text appears

The transcribed text is automatically inserted right where your cursor is.

New · On-device AI

Clean transcripts, on your Mac.

Optional: after transcribing, WhisperBar runs your text through a small on-device language model (Gemma 4) to fix grammar, strip filler words and respect self-corrections — all without ever sending your words to the cloud.

  • Understands self-corrections ("wait, I meant 9 o'clock" → "9 o'clock")
  • Removes "äh", "ähm", "uh", "um" without changing your voice
  • German and English out of the box, same on-device model
  • 100% offline — no API key, no usage fees, no tracking
  • New: optionally route Enhance through your local CLI — Claude Code, Codex or Gemini CLI — for near-instant results

Your voice stays private

WhisperBar was built from the ground up for maximum privacy. Everything happens locally on your Mac — no exceptions.

No cloud

Your audio recordings are never uploaded. Everything stays on your device.

No tracking

No telemetry, no analytics, no accounts. You stay anonymous.

No API keys

Everything runs locally — no external services, no extra costs.

Simple pricing

Pay once, use forever. No subscriptions, no hidden fees.

Full package
WhisperBar
€9.99
One-time purchase — no subscription
  • All 6 Whisper models
  • 10 languages supported
  • On-device AI enhancement (Gemma 4)
  • Smart text insertion via Accessibility API
  • Snippet expansion
  • Large v3 for maximum accuracy
  • All future updates included
  • 100% offline & private
  • Use on up to 3 devices simultaneously
Buy now for €9.99

Need a volume license for your team? Contact us for Enterprise pricing

10 languages supported

WhisperBar recognizes the most important languages — with the right model for each.

German English French Spanish Italian Portuguese Dutch Polish Japanese Chinese

Install via Homebrew

Prefer the terminal? Install and update WhisperBar with a single command.

brew install kchromik/apps/whisperbar

Frequently asked questions

No. You only need internet once to download the Whisper model. After that, WhisperBar works completely offline.

For German and other European languages we recommend "Small". For Japanese and Chinese, use "Medium" or larger. For maximum accuracy, choose "Large v3".

Yes. WhisperBar inserts text via the Accessibility API and works in any app with a text field — Notes, Mail, Safari, Slack, VS Code and more.

Enhance is an optional second step that runs a small on-device language model (Gemma 4) after transcription to tidy up grammar, punctuation and self-corrections. It runs 100% locally, you can turn it on or off at any time, and it never sends your text anywhere.

No. WhisperBar costs a one-time €9.99. No subscriptions, no in-app purchases, no hidden fees. All future updates are included.

WhisperBar requires macOS 15 (Sequoia) or later.

WhisperBar runs on all M-series Macs — M1, M2, M3, M4, M5, and future Apple Silicon chips. Intel Macs are not supported. We recommend at least 8 GB of RAM, as WhisperBar uses the larger Whisper Large V3 Turbo model for better accuracy.