Voice-Powered Document Search

Archivist lets you speak your questions instead of typing them. Ask a question out loud, and the AI searches your documents and responds — all running locally on your machine.

How It Works

On the Query tab, you'll see a voice input area with a microphone button:

1. Click the microphone button to start recording.
2. Speak your question naturally. A live waveform shows your audio as you speak.
3. Click the button again to stop.
4. Archivist transcribes your speech into text using a local speech-to-text model.

From there, the question is sent to the AI just like any typed message. It searches your documents and gives you an answer.

Auto-Send or Review First

You can choose what happens after your voice is transcribed:

• Auto-send on — Your question goes straight to the AI. Great for quick, conversational use when you're confident in the transcription.
• Auto-send off — The transcribed text appears in the input box so you can review, edit, or rephrase it before sending. Useful for technical terms or proper nouns that might not transcribe perfectly.

Completely Private

The voice recognition runs entirely on your computer using a local AI model. No audio is recorded to disk, sent to any server, or processed by any cloud service. Your voice data stays on your machine and is discarded after transcription.

When This Is Useful

• Hands-free research — Ask follow-up questions while reading through documents on screen.
• Accessibility — Interact with your documents without needing to type.
• Speed — Speaking a question is often faster than typing it, especially for longer or more detailed queries.
• Natural phrasing — People tend to ask more natural, conversational questions when speaking, which can sometimes produce better search results.