Handy
Free, open source desktop app that turns your speech into text on your own computer, offline.
- Stars
- 37k
- Forks
- 3.5k
- Licence
- MIT
- Language
- Rust
- Last update
- 4 days ago
- Type
- App
What it is
Handy is a cross-platform desktop app for speech-to-text. You press a keyboard shortcut, speak, and your words are typed or pasted into whatever text field is active. According to the README, transcription runs locally using Whisper or Parakeet V3 models, so audio is not sent to the cloud.
- Dictate emails, documents, or chat messages into any app by pressing a shortcut
- Control recording from Raycast using the community Raycast extension
- Start Handy hidden or without a tray icon on login using command-line flags
- Toggle recording from scripts with handy --toggle-transcription
Great for
- People who want dictation without sending their voice to a cloud service
- Anyone who wants a free, open source alternative to paid dictation tools
- People who prefer a simple hold-to-talk or toggle shortcut workflow
- Users on Windows, macOS, or Linux looking for offline transcription
- Accessibility-minded users who type by voice
Not ideal for
- Users on Wayland Linux who cannot install extra text input tools, since the README says Wayland support is limited and needs wtype or dotool
- Mac users who want the fn/Globe key as a shortcut on a third-party keyboard, because the README says it only works on Apple keyboards
- Anyone expecting a polished, fully stable product, since the README calls CLI flags a beta feature and the project is actively developed
How to get it
- 1
Download the app
Install the downloaded .deb on Debian/Ubuntu
sudo apt install ./Handy_*.debInstall on macOS via Homebrew cask (not maintained by Handy developers)
brew install --cask handyInstall on Windows via winget (not maintained by Handy developers)
winget install cjpais.Handy - 2
Run it
- Download the installer for your system from the releases page or handy.computer
- Launch Handy and grant microphone and accessibility permissions
- Set your preferred keyboard shortcut in Settings
- Press the shortcut, speak, and release to see your text appear
- 3
Get the most out of it
- Use the Hold-only, Toggle-only, or combined shortcut modes to match how you like to dictate
- Install xdotool on X11 or wtype/dotool on Wayland for more reliable text input
- Use command-line flags like --start-hidden and --no-tray for autostart setups
- Try the Raycast integration to control recording and browse transcript history
- Press Cmd+Shift+D (macOS) or Ctrl+Shift+D (Windows/Linux) to open debug mode when troubleshooting
Downloads
Strengths
- Runs locally, so the README says audio stays on your computer
- Free and MIT licensed, with source code available to extend
- Lets you choose between several Whisper model sizes and Parakeet V3
Trade-offs
- Linux text input needs extra tools such as xdotool, wtype, or dotool, and Wayland support is described as limited
- Some features are beta or have known issues, such as Bluetooth headset audio quality on macOS
- Model choice and GPU use affect speed, and the README does not give performance numbers
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| Handy (this) | 37k | App | beginner | 4 days ago | People who want dictation without sending their voice to a cloud service |
| whisper.cpp | 54.9k | Library | intermediate | 7 days ago | Developers who want to embed offline speech-to-text in a C, C++, or other native app |
| buzz | 22.4k | App | beginner | 4 days ago | People who want to transcribe recordings without uploading them to an online service |
| openwhispr | 10.5k | App | beginner | 4 days ago | People who want dictation in any app without sending audio to a cloud service |
| NextChat | 88.9k | App | beginner | 17 days ago | People who want one chat window for several AI models |
| micro | 29.8k | App | beginner | 4 days ago | People who work mostly in a terminal and want an easy editor |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.