Alternatives / Superwhisper
Looking for an open source Superwhisper alternative? We compared 5 open source voice dictation apps by license, platforms, local transcription, AI cleanup, and pricing. They transcribe your speech on your own device and type it into any app, and two of them sell paid licenses even though the code is open.
Written by Maxime DERAME
Last updated:
| Alternative | Best for | License |
|---|---|---|
| VoiceInk | Closest macOS dictation app (paid license, free to build) | GPL-3.0 (paid builds) |
| Handy | Simple cross-platform local dictation | MIT |
| FluidVoice | Free macOS dictation with on-device AI polish | GPL-3.0 |
| Amical | Context-aware dictation with local or cloud models | MIT |
| VoiceTypr | Mac and Windows dictation with a lifetime license | AGPL-3.0 (lifetime license for builds) |
We looked for open-source tools that overlap with Superwhisper's core: pressing a hotkey, speaking, and getting accurate text in any app, with local transcription for privacy. We verified each project's license, recent activity, stars, and platforms in its repository. We left out two tools listed elsewhere because we could not find a working repository or website for them. See our methodology for the full criteria.
VoiceInk publishes its source under GPL-3.0 but sells a paid license with a free trial, and asks users who do not build it themselves to purchase. VoiceTypr is AGPL-3.0 with a trial and a lifetime license for its built binaries. You can build either from source, but the polished downloads are paid.
VoiceInk is the closest: a macOS app with local transcription, optional AI cleanup, and per-app settings. FluidVoice is a free macOS alternative with on-device AI polish, and Handy is the simplest cross-platform option, available on Windows, macOS, and Linux with a deliberately minimal feature set. Amical adds context-aware formatting with local or cloud processing.
VoiceInk needs Apple Silicon and a recent macOS, and FluidVoice runs on macOS today with iOS and Windows planned. Handy runs on Windows, macOS, and Linux, and VoiceTypr runs on macOS and Windows. Local models need enough memory and compute, and an optional FluidVoice model is several gigabytes.
VoiceInk and FluidVoice are GPL-3.0, VoiceTypr is AGPL-3.0, and Handy and Amical are MIT. FluidVoice also uses a separate, privately maintained local AI runtime for its advanced enhancement.
All five projects were updated within weeks of this update. Amical's last commit was in September 2026.
Superwhisper is a proprietary macOS app with a subscription or paid license, and its best models and features sit behind paid tiers. Open-source alternatives run speech recognition locally with models such as Whisper and Parakeet, keep your audio on your device, work on more platforms in some cases, and let you inspect the code, at the cost of a less polished experience and your own hardware requirements.
Handy is an MIT-licensed desktop app for Windows, macOS, and Linux that types your speech into any text field, with transcription processed locally. It has a deliberately minimal feature set with few settings and can be installed with Homebrew or winget.
Pricing: Free and open source.
FluidVoice is a free, GPL-3.0 macOS dictation app with on-device speech recognition and an AI polish layer. Its advanced enhancement uses a separate, privately maintained local AI runtime, and post-processing can optionally route to OpenAI, Groq, or a custom provider. iOS and Windows versions are planned.
Pricing: Free and open source; optional providers are billed separately.
VoiceInk is a macOS dictation app that transcribes your speech locally into any text field, with optional AI cleanup and per-app settings. The source code is GPL-3.0, but the polished builds are sold with a free trial and a paid license, and building it yourself is possible. It needs Apple Silicon and a recent macOS.
Pricing: Free to build from source; the prebuilt app is a paid license with a free trial.
Amical is an MIT-licensed AI dictation app with context-aware formatting, powered by open models that work offline, with local or cloud processing. It aims to let you type faster without a keyboard.
Pricing: Free and open source.
VoiceTypr is an AGPL-3.0, offline-first dictation app for macOS and Windows that uses Whisper models and Apple Silicon-optimized Parakeet models on macOS. After a trial, a lifetime license keeps local models unmetered with no subscription, and optional cloud modes use your selected provider's billing.
Pricing: Free trial, then a lifetime license; building from source is possible.
| Alternative | Platforms | License | Local transcription | Best for |
|---|---|---|---|---|
| VoiceInk | macOS | GPL-3.0 (paid builds) | Yes | macOS dictation |
| Handy | Windows, macOS, Linux | MIT | Yes | Simple dictation |
| FluidVoice | macOS | GPL-3.0 | Yes | Free macOS dictation |
| Amical | Desktop | MIT | Yes | Context-aware formatting |
| VoiceTypr | macOS, Windows | AGPL-3.0 (lifetime license) | Yes | Mac and Windows |
Want the closest macOS dictation app
VoiceInk offers local transcription, optional AI cleanup, and per-app settings, but the built app is a paid license.
Want free local dictation on a Mac
FluidVoice is free under GPL-3.0 with on-device transcription and AI polish, though it runs on macOS only.
Want simple dictation on any desktop
Handy is an MIT app that types speech into any field on Windows, macOS, and Linux, processed locally.
Want formatting that adapts to the app you use
Amical is an MIT dictation app with context-aware formatting and local or cloud processing.
Want a one-time purchase on Mac and Windows
VoiceTypr is AGPL-3.0 with on-device models and a lifetime license instead of a subscription.