AudioWhisper
A macOS dictation app that transcribes audio offline with WhisperKit or Parakeet-MLX, with optional OpenAI Whisper and Google Gemini processing.
AudioWhisper is a macOS menu bar app for dictating into other apps and converting recordings to text. It suits people who want on-device speech recognition with the option to use cloud transcription.
- Dictation across apps: Global shortcuts, push-to-talk and a windowless recording mode let you record while working elsewhere, then copy or automatically paste the transcript and restore focus to your previous app.
- Local transcription: WhisperKit runs Whisper models through CoreML, while Parakeet-MLX provides a multilingual alternative; both work offline after model downloads and need no API key.
- Text cleanup: Optional correction uses local MLX or a cloud provider to fix punctuation and typos and remove filler words, with editable prompts for contexts such as coding and email.
- Audio files and history: You can transcribe existing recordings and optionally save searchable transcripts locally with retention controls; a usage dashboard tracks words, dictation speed and estimated typing time saved.
The app requires macOS Sonoma or later and is open source under the MIT license. Local Whisper works on Intel Macs but runs more slowly; Parakeet-MLX and local MLX correction require Apple Silicon.
Local transcription keeps audio on your Mac. OpenAI Whisper and Google Gemini require internet access and API keys and receive audio for processing. Cloud correction also sends text to the selected provider. AudioWhisper stores API keys in macOS Keychain and collects no usage analytics.