Introduction to VoiceInk

VoiceInk turns speech into text using a local model or a cloud provider you connect. A Mode chooses the transcription model, language, optional AI enhancement, and how the result is delivered.

Start Dictating

1. Install VoiceInk and complete onboarding.
2. Download a local transcription model or connect a cloud provider in AI Models.
3. Choose the model in a Mode.
4. Click the destination text field, then use your recording shortcut to start and stop recording.

A Mode with Paste output inserts the finished text at your cursor. Respond shows an AI answer in the recorder. Custom Command passes the result to a local shell command.

Features

• Modes and triggers: Choose different workflows for apps, websites, spoken trigger phrases, and shortcuts.
• Local and cloud models: Use on-device transcription or your own provider API keys.
• Real-time transcription: Process supported models while you speak and optionally display partial text.
• AI enhancement: Clean up dictation, rewrite text, or draft replies with your chosen provider and prompt.
• VoiceInk Refine: Clean up dictation locally on a supported Mac.
• Context awareness: Include selected text, clipboard text, or text extracted from the active window in enhancement requests.
• Dictionary: Add vocabulary, apply word replacements, and learn useful corrections.
• Assistant Mode: Ask questions and continue a conversation inside the recorder.
• File transcription: Process existing audio and video files in a queue.
• History and Quick History: Search, replay, copy, paste, export, and retry saved results.
• Dashboard and Insights: View activity, estimated time saved, and model performance.
• Backups: Export supported settings and restore selected categories.

Local transcription keeps speech processing on your Mac. Dictionary Auto Learn defaults to local rules, but optional AI review and enhancement can use cloud providers. A local transcription model alone does not make every workflow offline. See Privacy and Data.