Wren
Back to Wren alternatives

Wren vs Handy

An open-source, offline-only dictation app for macOS, Windows, and Linux, built on Tauri and local Whisper/Parakeet models.

At a glance

Platforms
macOS, Windows, Linux
Open source
Yes — view source
Free tier
Free, MIT licensed, no paid tier. Development is funded through GitHub Sponsors and donations, not subscriptions.

What Handy is good at

Handy is a mature, widely used open-source dictation app: press a shortcut, speak, and it pastes the transcript into whatever app has focus, entirely offline. It runs Whisper (Small through Large, with GPU acceleration where available) or Parakeet V3 for CPU-only setups, ships official signed and notarized builds for macOS, Windows, and Linux, and has been translated into over twenty languages by its community. If you want local dictation and never need a cloud option, Handy already does that well.

How Wren is different

  • Wren can optionally send audio to a cloud provider you configure yourself (bring your own API key, including free options like Groq); Handy has no cloud or remote transcription path at all, by design, so it's offline-only in every configuration.
  • Wren renders its recording indicator as a native wgpu overlay with no webview involved; Handy's overlay is a Tauri window, which is why Handy disables it by default on Linux and documents that some compositors treat it as the active window and can steal focus.
  • Handy currently supports more offline models than Wren: Whisper at four sizes plus Parakeet V3, versus Wren's single embedded model (Parakeet V3, multilingual, CPU) today.
  • Handy ships signed, notarized builds for macOS and Windows through its release pipeline (Apple Developer ID plus notarization, Azure Trusted Signing); Wren's builds aren't code-signed yet, so Windows SmartScreen and macOS Gatekeeper currently flag them as from an unrecognized developer, a limitation disclosed in Wren's own release notes.
  • Handy's UI has been translated into over twenty languages by community contributors; Wren currently ships English and Portuguese (pt-BR) only.
  • Both are MIT-licensed, Tauri-based, Rust-plus-React desktop apps with a global shortcut, VAD-based silence trimming, and paste-based delivery. Wren's architecture for the embedded/offline engine took cues from Handy's approach, so on the core offline dictation loop the two apps are close cousins.

Questions, answered

Is Handy really free, with no catch?

Yes. Handy is MIT licensed, has no paid tier, and its README states the project's own position: accessibility tooling shouldn't sit behind a paywall. Development is supported through GitHub Sponsors, Ko-fi, Buy Me a Coffee, and PayPal donations rather than subscriptions.

Does Handy support cloud transcription?

No. Handy transcribes entirely on-device using local Whisper or Parakeet models; its own documentation is explicit that no audio or information is sent to the cloud. Wren also transcribes offline by default, but you can opt into a cloud provider with your own API key.

Is Handy or Wren more mature right now?

Handy is the more established project as of this review: it has a much larger user base, more offline model options (Whisper at multiple sizes plus Parakeet), broader UI translation, and signed, notarized release builds on all three platforms. Wren is newer, ships one embedded model, and its builds aren't code-signed yet. Wren's main practical difference is the optional bring-your-own-key cloud path alongside offline transcription.

Is Wren a fork of Handy?

No, they're separate codebases. Wren's embedded/offline transcription engine drew architectural ideas from Handy's approach, but Wren is written independently and adds a provider-agnostic layer (cloud, local server, or embedded, chosen at runtime) that Handy doesn't have.

Sources

Pricing and features last checked 2026-09-09. Details change — verify against the source before relying on them.

Back to Wren alternatives