Otter.ai
AI meeting assistant for transcription and automated note-taking
Handy
Offline, privacy‑first speech‑to‑text app for all platforms
Handy offers free offline speech transcription for Windows, macOS, and Linux. Use a shortcut, pick Whisper or Parakeet, and paste text into any app privately.
Overview
Overview
Handy is a cross‑platform desktop application that brings on‑device speech transcription to Windows, macOS, and Linux. It targets users who need a reliable dictation tool without sacrificing privacy, as all audio processing stays on the local machine.
Capabilities
Activate recording with a configurable global shortcut, speak, and let Handy handle silence filtering via Silero VAD. Transcriptions are generated by either OpenAI Whisper models (with optional GPU acceleration) or the CPU‑optimized Parakeet V3 model, which also auto‑detects language. The resulting text is automatically pasted into the active window, making it easy to insert into any editor, browser, or communication app. Built with Tauri, Rust, and React/TypeScript, the codebase is designed for extensibility, allowing developers to fork and integrate the engine into their own projects.
Deployment
Download the pre‑built binaries from the releases page, grant microphone and accessibility permissions, and configure shortcuts in Settings. No external services or internet connection are required, ensuring a fully offline experience.
Highlights
Local transcription using Whisper and Parakeet models
Configurable global keyboard shortcuts for start/stop
Silero Voice Activity Detection filters silence
Cross‑platform UI built with React and Tauri
Pros
- Fully offline operation preserves user privacy
- GPU acceleration available for Whisper models
- CPU‑optimized Parakeet runs on low‑end hardware
- Open‑source and easily extensible
Considerations
- Optimal Whisper performance needs modern GPU
- Language support limited to model capabilities
- No built‑in cloud backup or synchronization
- Debug mode may be confusing for non‑technical users
Managed products teams compare with
When teams consider Handy, these hosted platforms usually appear on the same shortlist.
Looking for a hosted option? These are the services engineering teams benchmark against before choosing open source.
Fit guide
Great for
- Users needing private, on‑device speech transcription
- Developers who want to extend an open‑source speech tool
- Professionals working across Windows, macOS, and Linux
- Accessibility advocates seeking free dictation software
Not ideal when
- Environments without microphone access
- Users expecting real‑time network streaming transcription
- Low‑end devices lacking required CPU/GPU for Whisper
- People needing built‑in cloud services or mobile apps
How teams use it
Instant dictation in word processors
Press the shortcut, speak, and have text appear directly in Word, Google Docs, or any editor without leaving the document.
Transcribing meeting notes
Capture spoken discussion, then paste the transcript into note‑taking apps like Notion or Obsidian for quick reference.
Accessibility aid for limited typing
Provide a reliable, offline voice input method for any application, enhancing productivity without compromising privacy.
Custom speech‑to‑text integration
Fork Handy to embed its transcription engine into other Rust or JavaScript projects, leveraging the same local models.
Tech snapshot
TypeScript54%
Rust45%
CSS1%
HTML1%
JavaScript1%
Frequently asked questions
Does Handy require an internet connection?
No. All audio processing and model inference happen locally on your computer.
Which speech models are supported?
Handy can run OpenAI Whisper models (Small, Medium, Turbo, Large) with optional GPU acceleration, and the CPU‑optimized Parakeet V3 model.
How do I set the hotkey?
Open Settings, navigate to Keyboard Shortcuts, and assign any key combination for start/stop or push‑to‑talk mode.
Can I use Handy on a Mac with Apple Silicon?
Yes. Handy runs on both Intel and Apple Silicon Macs, leveraging GPU acceleration when available.
Is there a way to contribute?
Yes. Fork the repository, make changes, and submit a pull request; see the CONTRIBUTING guidelines on GitHub.
Project at a glance
Active- Stars
- 12,559
- Watchers
- 12,559
- Forks
- 836
LicenseMIT
Repo age11 months old
Last commit2 days ago
Primary languageTypeScript
Last synced 12 hours ago