Handy

Free, open-source push-to-talk speech-to-text for Windows, macOS and Linux that runs entirely offline.

Visit website (opens in a new tab) Compare with other tools

Handy is an MIT-licensed desktop app by CJ Pais that pastes what you say into any text field. Press or hold a shortcut, speak, release; audio is transcribed on your machine with a downloadable Parakeet, Whisper, Moonshine, SenseVoice, GigaAM or Canary model and nothing is sent to the cloud. It keeps a local history with audio playback, supports custom words and has an experimental post-processing step that can send the transcript to a cloud or local LLM for cleanup.

Pros

  • Completely free and open source (MIT, 31k+ GitHub stars); no account, subscription or cloud service
  • Interface available in 18+ languages, including English, Chinese, Japanese, Korean, French and Spanish
  • Choice of local models such as Parakeet V3 (25 European languages), Whisper (99+ languages) and SenseVoice (Chinese, Japanese, Korean, Cantonese)
  • Windows, macOS and Linux builds including ARM, plus CLI flags and signal control for Wayland and tiling window managers

Cons

  • No live captions; text appears only after you stop recording
  • AI cleanup is an alpha feature that needs your own API key or a local LLM server
  • Spoken punctuation commands are not supported and recordings over 3-5 minutes may fail
Platforms
  • macOS
  • Windows
  • Linux
Speech processing
On-device
Works offline
Yes
Languages
99+

Pricing

Pricing model
Open source
Free tier
Yes

Features

  • Yes Works in any app
  • No Real-time transcription
  • Yes Auto punctuation
  • Yes AI cleanup
  • Yes Custom vocabulary
  • No Voice commands
  • Yes Push to talk
  • Yes Auto language detection
  • Yes Translation
  • No File transcription
  • No Meeting notes
  • Yes Multilingual interface
  • Yes History

Compare Handy with

Last verified 2026-09-19