2.1k

Ghost Pepper

Voice dictation and meeting transcription
without data ever leaving your machine.

100% local models = 100% privacy

Download for Mac

macOS 14.0+ · Apple Silicon (M1+) · Free & open source

★ Star on GitHub to support the project

MIT License 50+ languages 100% local No account required

Speech-to-text

Hold Control to talk, release to transcribe and paste into any text field. Works everywhere.

Meeting transcription

Record calls with notes, transcript, and AI-generated summaries — saved as local markdown.

Smart cleanup

Local LLM removes filler words, fixes self-corrections, and cleans up your speech automatically.

Completely private

All models run on your Mac via Apple Silicon. Nothing is uploaded, tracked, or stored in the cloud.


What people are saying

Ryan Hoover — Product Hunt
Dave Morin tweet Nick Saltarelli tweet
Michael Mignano tweet

Getting started

  1. Download GhostPepper.dmg
  2. Open the DMG, drag Ghost Pepper to Applications
  3. Grant Microphone and Accessibility permissions when prompted
  4. Hold Control and speak
"Apple could not verify" warning? Go to System Settings > Privacy & Security, scroll down, and click Open Anyway. You only need to do this once.

Privacy audit

Every core feature runs 100% on your Mac — verified by AI code review.

FeatureStatusWhat was checked
Speech-to-text LocalWhisperKit/FluidAudio inference, no audio sent anywhere
Text cleanup LocalQwen LLM runs on-device via LLM.swift
Meeting transcription LocalChunked transcription, markdown files on disk
File storage LocalMarkdown to local filesystem, no cloud sync
Analytics & telemetry NoneNo tracking SDKs of any kind

Models

Speech ModelSizeBest for
Whisper tiny.en~75 MBFastest, English only
Whisper small.en (default)~466 MBBest accuracy, English only
Parakeet v3~1.4 GB25 languages
Qwen3-ASR 0.6B~900 MB50+ languages, macOS 15+
Cleanup ModelSizeSpeed
Qwen 3.5 0.8B (default)~535 MBVery fast (~1-2s)
Qwen 3.5 2B~1.3 GBFast (~4-5s)
Qwen 3.5 4B~2.8 GBFull quality (~5-7s)