Handy is a free, open-source dictation app for macOS, Windows, and Linux that turns speech into typed text anywhere you can type. You press a global shortcut, speak, and release: the app transcribes locally with Whisper or Parakeet models, trims silence with voice activity detection, and types the result into the focused field.[1][2]
Quick verdict
Desktop users who dictate across many apps and need voice data to stay on the machine get the most from Handy: local Whisper and Parakeet transcription plus automatic silence trimming turns speech into finished text in any field, free of charge. The decisive tradeoff is platform polish and setup: documented rough edges on Bluetooth microphones, the macOS fn key, Linux Wayland, and overlay windows, plus multi-hundred-megabyte model downloads, mean anyone wanting a zero-friction hosted service should choose the Wispr Flow incumbent instead.[1][2]
Key features
Global-shortcut dictation anywhere Global-shortcut dictation anywhere: a system-wide hotkey starts recording over any application, and the transcript is typed into the focused text field or copied to the clipboard.[1][2]
Fully offline Whisper and Parakeet transcription Fully offline Whisper and Parakeet transcription: speech is converted on the device by downloadable local models, including Whisper variants and Parakeet V3, so dictation audio never leaves the machine.[1][2]
Automatic silence handling Automatic silence handling: bundled Silero voice activity detection trims quiet stretches and stops recording automatically, keeping transcripts free of long pauses.[1][2]
GPU acceleration on three platforms GPU acceleration on three platforms: Metal on Apple Silicon, CUDA on NVIDIA, and Vulkan elsewhere shorten transcription turnaround, with CPU-only fallback supported.[1][2]
99 languages with automatic detection 99 languages with automatic detection: the transcription language is detected from speech, and models install per language, covering multilingual users without manual switching.[1][2]
Two recording styles plus history Two recording styles plus history: press-and-release dictation suits quick insertions, hold mode suits longer passages, and past transcriptions stay reviewable in settings.[1][2]
Dictionary, extensions, and automation hooks Dictionary, extensions, and automation hooks: a custom dictionary teaches names and jargon, community extensions adjust per-app behavior, and CLI flags plus a debug mode support scripted use.[1][2]
Three-platform installers and open source Three-platform installers and open source: signed installers cover macOS, Windows, and Linux, the MIT-licensed codebase builds from source, and a Raycast extension adds history, dictionary, and model switching.[1][2]
Who will love it
Privacy-conscious dictators of sensitive material: on-device transcription means spoken words stay off hosted services entirely.[1][2]
Cross-platform users who refuse subscriptions: one free MIT-licensed app covers macOS, Windows, and Linux with no recurring fees.[1][2]
Multilingual speakers and jargon-heavy writers: automatic language detection across 99 languages plus a custom dictionary keeps specialized dictation spelled correctly.[1][2]
Who should avoid it
Users wanting a zero-setup hosted service: local models must be downloaded at roughly 500 MB to several GB, which is more setup than a cloud dictation subscription.[1][2]
People on affected hardware and windowing setups: documented issues cover macOS Bluetooth microphones and the fn key, Linux Wayland and some text tools, overlay windows, and clipboard paste in certain apps.[1][2]
Mobile-first dictators: the roadmap lists mobile support as planned, but current releases are desktop-only, so phone-first dictation has no supported path.[1][2]
Alternatives and compatibility
Wispr Flow is the incumbent in the published catalogue relationship and the honest hosted redirect: a managed dictation service for users who prefer convenience over locality.[1][2]
Handy's decisive difference is locality plus license: transcription runs fully offline with no account under an MIT codebase, which is precisely the property that sends privacy-conscious users looking for a local alternative.[1][2]
Choose Handy when dictation must stay on the machine, span three desktop platforms, and cost nothing; choose the hosted incumbent when cross-device sync and zero local setup matter more than locality.[1][2]
Questions worth asking
What is the best Handy alternative for non-technical users?
Is Handy really no-code?
How much does Handy cost per month?
Can you self-host Handy?
There is nothing to host: Handy is an MIT-licensed open-source desktop app that runs entirely on your machine, and you can clone the repository and build it yourself on macOS, Windows, or Linux rather than downloading a release. Self-operation means keeping your own build working on supported platforms.[1]
Sources and editorial review
Source research. Reviewed on October 11, 2026. Written by Sanjib Sah. AI assisted; reviewed by a human editor.
We have not run or tested this product.
- Handy official site (handy.computer)
official · Observed October 11, 2026
Official Handy site: free open-source offline speech-to-text, global-shortcut dictation into any app, extensible platform, roadmap (Linux, MCP, mobile), FAQ replacement positioning against Superwhisper/WisprFlow with local Whisper, competitor pricing comparison, sponsorship tiers $3-$3,000/mo.
- Handy repository README (official)
official · Observed October 11, 2026
Official Handy GitHub repository (cjpais/Handy): offline-first dictation, automatic language detection, Silero VAD, Whisper plus Parakeet V3 models, GPU acceleration (Metal, CUDA, Vulkan), hold/release modes, 99 languages, macOS/Windows/Linux installers and source builds, Raycast extension, CLI flags, clipboard-output mode, known platform issues, 33.5k stars, MIT license with brand-use restriction, created February 2025.