Local dictation for Mac: speak messy, send clean
BetterWispr is a free, open source dictation app for macOS. Hold a shortcut, speak, and it transcribes locally on your Mac, cleans filler words, and types styled text into whatever app you were using. No account is required, nothing is uploaded, and meeting notes are built in.
Built-in models transcribe speech entirely on your Mac with no account and no cloud fallback. Nine built-in models including Apple speech, Parakeet TDT v3 (25 European languages) and Parakeet Japanese, Whisper Large v3 Turbo via WhisperKit, or your own Sarvam AI, Smallest AI or OpenAI-compatible API
Hold Option + Space while speaking and release to transcribe; text lands directly in the app you were using. Supports press-to-toggle and custom shortcuts (⌘, ⌥, ⌃ or an F key), with clipboard fallback when paste is blocked.
Set a tone for personal chats, work apps and email: casual drops the final period, very casual also drops capitals, formal keeps words as spoken. Styles apply to English.
Light cleanup (default) removes English filler words, repeats and false starts; Medium also edits for clarity using Apple Intelligence or Ollama on your Mac; None keeps every word.
Say comma, question mark, or new paragraph while dictating; say 'scratch that' to delete the last sentence (English only).
Add names and terms as hints and replacements in Vocabulary, or fix a misheard word in History or right after a paste — it spells your way next time, and learning can be turned off.
Transcribes your microphone as Me and (on macOS 14.2+) other apps' audio as Them, next to your own notes. Apple Intelligence or a local Ollama model writes summaries, decisions and action items on your Mac.
Dictations stay on your Mac with raw and corrected text — 'Use Original' restores exactly what you said, and history can be turned off. Insights show words spoken, pace, words cleanup saved, apps dictated into and daily streaks, all computed locally.
Install BetterWispr and follow the welcome guide through permissions and your first local speech model.
Place your cursor in a text field, hold Option + Space, and say what you want to write.
Let go to transcribe; clean text arrives in the app you were using, in the style you set for it.
BetterWispr is free and open source under the Apache 2.0 license. There is no dictation subscription, account or API key requirement; model files use storage on your Mac. Optional API providers (Sarvam AI, Smallest AI, OpenAI-compatible endpoints) are bring-your-own-key.
Free
Full open source dictation app under Apache 2.0 — no subscription, account or API key required. Optional third-party API providers can be connected with your own key.
A free, open source dictation app for macOS. You hold a shortcut, speak, and BetterWispr transcribes locally on your Mac, cleans up filler words, and types styled text into whatever app you were using — with no account and no cloud fallback.
It's free. BetterWispr is open source under Apache 2.0 with no dictation subscription, account or API key requirement. Model files use storage on your Mac.
Yes, once a compatible local model and its assets are installed. Downloads happen when you choose to install them, and there is no cloud transcription fallback.
It depends on the model: Parakeet TDT v3 covers 25 European languages, Parakeet Japanese covers Japanese, and Whisper offers broader multilingual recognition. Filler cleanup, styles and voice commands work in English.
Any Mac on macOS 14 or later, with Apple Silicon recommended for local model inference. Model availability and performance depend on your hardware and language.
Focus a text field, hold Option + Space, speak, and release to transcribe. You can switch to press-to-toggle or set a different shortcut in Settings. Transcription begins after recording stops rather than appearing live while you talk.
Only as much as you choose. Cleanup starts at Light, which removes English filler words, repeats and false starts. Medium also edits for clarity using Apple Intelligence or Ollama, and None keeps every word. History keeps the raw transcript so you can always restore the original.
Yes. Notetaker transcribes your mic as Me and other apps' audio as Them (macOS 14.2+) while you write your own notes, then Apple Intelligence (macOS 26) or a local Ollama model writes the summary, decisions and action items on your Mac.
