OpenSuperWhisper is a free, open-source dictation app for macOS that turns your voice into text in any application — entirely on your own Mac. Inspired by premium dictation tools, it runs state-of-the-art speech recognition locally using NVIDIA’s Parakeet and OpenAI’s Whisper model families, so your audio never leaves the machine. It has attracted about 3,100 stars on GitHub and is actively maintained, with updates in 2026.
The workflow is built for speed: hold a hotkey (or toggle recording), speak, and the transcribed text is pasted at your cursor — in email, chat, code editors, browsers, anywhere you can type. Because transcription runs locally on Apple Silicon through optimized model runtimes, results arrive quickly and keep working offline, and you can choose between Parakeet models for near-instant English/European-language dictation or Whisper models for broader language coverage. The app keeps a recording history so you can recover or re-copy past dictations, and settings cover model choice, hotkeys and audio input. For writers, developers and heavy message-senders who think faster than they type, OpenSuperWhisper provides the SuperWhisper-style experience as a transparent, hackable open-source project.
OpenSuperWhisper is completely free and open source under the MIT license. There are no usage limits or cloud fees because recognition runs on your own hardware.
OpenSuperWhisper is for Mac users — writers, developers, support staff and anyone with tired hands — who want fast, private voice typing everywhere on their computer without paying for a dictation subscription.