Press a hotkey, speak naturally, and get polished text in any app. VoxType transcribes your speech locally, then refines it with the AI model of your choice.

Trigger VoxType from any app with a system-wide keyboard shortcut. The floating panel appears without stealing focus from your current window.
VoxType transcribes your speech in real-time using on-device Apple Speech or the Whisper model. Your audio never leaves your Mac.
Choose a writing style to polish your text, translate it, or insert it as-is. The result is placed directly into your active app.
Activate VoxType from any application with a single keyboard shortcut. No need to switch windows.
The floating transcription panel appears without taking focus away from your current app. Your cursor stays where it is.
Choose between on-device Apple Speech Recognition for speed or the Whisper model for accuracy. Both work locally on your Mac.
Apply predefined or custom writing styles to refine your transcribed text. Turn casual speech into professional prose.
Connect your own API keys for OpenAI, Anthropic Claude, or Google Gemini. Use the AI model you prefer.
Speak in one language and get polished text in another. VoxType can translate your speech as part of the refinement step.
Review past transcriptions and refined outputs. Copy or re-use previous results without re-recording.
Define per-app or per-website rules for writing style and language. VoxType adapts its behavior automatically.
Audio is processed locally on your device. Only text is sent to AI models when you choose to refine. No audio is ever uploaded.
Common questions about VoxType