R_REDDYX.XYZ
VoiceRust★ 6,802⑂ 424 forksMIT

anomalyco/hex

LOCAL VOICE DICTATION FOR ALL OS

Hex is a local voice dictation app for macOS, Windows, and Linux. Hold a key, speak, and the text pastes into the active window.

// KEY FEATURES

  • Works offline after model download
  • Supports macOS, Windows, Linux
  • Transcribes audio/video files via CLI
  • Written in Rust, distributed under MIT

// DEEP DIVE: HOW IT WORKS

Hex works on a simple principle: the user holds a configured key, speaks, releases — and the recognized text is automatically pasted into the current application. On first launch, a speech model is downloaded, after which all processing happens locally; no data leaves the device. The app is written in Rust and shipped as native binaries for all major platforms: macOS (Apple Silicon), Windows (x64 and ARM64), Linux (x86_64 and ARM64) with .deb, .rpm packages and universal executables.

The distinguishing feature of Hex is full locality and ease of system integration. Unlike cloud services such as Google Voice Typing or Whisper API, there are no network requests after the initial model download. It includes a full-featured CLI (`hex transcribe`, `hex dictation`, `hex settings`) allowing transcription of existing audio and video files (WAV, MP3, MP4, WebM, etc.) and controlling the app from scripts. Experimental features live in Labs. The project does not accept pull requests: bugs and ideas are submitted via Issues, which the team resolves quickly.

Requirements: macOS on Apple Silicon, Windows 10/11 x64 or ARM64, Linux x86_64 or ARM64 (glibc). A microphone and permissions to input text into other apps are needed (on macOS — Accessibility, on Linux — appropriate Wayland/X11 permissions). The model downloads on first run; its size is not specified in the README. MIT licensed, source code is open, but the development process is closed to external contributors.

// USE CASES

Best for: For those who want voice input without sending data to the cloud

  • Dictate text in any editor by holding a hotkey
  • Transcribe meeting recordings and videos via hex transcribe command
  • Control voice input from scripts and automate routine

// QUICK START

brew install --cask anomalyco/tap/hex
yay -S hex-bin
curl -fsSL https://anoma.ly/hex/install-linux.sh | sh
nix profile add github:anomalyco/hex

Commands are taken from the repository README — check the current version on GitHub before running.

#dictation#linux#local-first#macos#rust#speech-recognition#speech-to-text#voice-control
Open on GitHub →

New repositories every 30 minutes

REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.

Join on Telegram
← Full catalog·Full index

// SIMILAR REPOSITORIES

DOCKER WRAPPER KOKORO-82M TTS WITH OPENAI API★ 5706VIETNAMESE TTS WITH ON-DEVICE INSTANT VOICE CLONING★ 3432SMALL ENGLISH TTS MODEL 4.5M PARAMETERS★ 673LOCAL SPEECH-TO-SPEECH ASSISTANT ON GPU★ 448OFFLINE AI VOICE STUDIO IN RUST★ 312VOICE STUDIO ON MAC WITH MLX★ 247

← FULL CATALOG

anomalyco/hex — Local Voice Dictation For All OS | REDDYX AI