Handy
A free, offline push-to-talk dictation app for macOS, Windows and Linux that runs Whisper or NVIDIA Parakeet on your own machine and pastes the text at your cursor, under MIT.
What it does
Handy describes itself as "a free, open source, and extensible speech-to-text application that works completely offline". You hold a key to record and release it to stop, or tap the key to toggle. The text is pasted into whatever field is active.
The speech model runs locally. The README lists Whisper in Small, Medium, Turbo and Large sizes with GPU acceleration where available, and NVIDIA's Parakeet V2 and V3, which it describes as CPU-optimised with automatic language detection. Custom Whisper models dropped into the models folder are picked up. Silence is trimmed with Silero VAD before transcription. A command-line flag toggles a post-processing step; the README does not say what that step does.
What it replaces
Handy replaces Wispr Flow and Superwhisper for the core job: speak, get text where the cursor is, with no subscription and no audio leaving the machine. Its README says "Your voice stays on your computer."
Where it falls short: there is no language-model cleanup you can configure from the app, no meeting capture, no phone version, and the README's own summary is "One tool, one job." The maintainer writes that "Handy isn't trying to be the best speech-to-text app."
How it runs
Installers on the releases page and the website, plus a Homebrew cask, a winget package and a .deb for Debian and Ubuntu; the README notes the Homebrew and winget packages are community-maintained. It is a Tauri app written in React and Rust, buildable from source.
For Whisper it wants an Apple Silicon or Intel Mac, or a discrete GPU on Windows or Linux. Parakeet V3 runs on the CPU with an Intel 6th generation Skylake or equivalent AMD as the floor, and the README quotes about five times real-time speed on a mid-range i5.
Who it fits
Anyone who wants a dictation key on a laptop and does not want an account. Someone who wants polished prose from rough speech should look at VoiceInk or OpenWhispr, which add a language-model step.
Limits the README states
Whisper models "crash on certain system configurations (Windows and Linux)". Wayland support is limited. The fn and Globe key shortcuts work only on Apple keyboards. Bluetooth headset microphones on macOS may reduce playback quality while recording. The recording overlay can interfere with pasting on Linux under X11.
Sources
- Handy README and LICENSE: github.com/cjpais/Handy, read 2026-09-16. The README notes the Handy name, logo and brand assets are excluded from the MIT grant.
More voice dictation apps
A dictation and meeting-notes app under MIT that runs Whisper or Parakeet locally or cloud models with your own key, detects Zoom, Teams and FaceTime calls, and labels speakers on device.
A macOS dictation app under GPL-3.0 whose README names Superwhisper and Wispr Flow as the tools it replaces, with local models, per-app modes, a personal dictionary and an AI assistant mode.
A speech-to-text app inside the Epicenter monorepo under AGPL-3.0 that records, transcribes with a provider you choose, cloud or local, optionally polishes the text, and pastes it at the cursor.
Added September 16, 2026. Found an error on this page? Write to reveneau@licheo.com and it gets fixed in the next weekly pass.
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.