Better context for your AI.

Open-source voice dictation with the screen context you point at.

Apple Silicon · macOS 14+ · local transcription

BetterVoice Settings overview showing dictation, visual context, and customizable shortcuts
Imagine Wispr Flow, but with visual context built in.

See it in action

Voice plus visual context.

The useful part

Your words need a picture.

Point at the part of the screen your agent needs to see.

Speak naturally

Hold Option for a quick note. Use Command–Option for more.

Point at what matters

Circle a button, error, or layout. The full display comes with it.

Paste it together

Transcript first. Images next. Everything goes to the field you were using.

Built for real workflows

Private by default.

Speech runs locally. Temporary audio is deleted. Captured sessions stay on your Mac and expire automatically.

Read the privacy and storage details
VoiceRun the server and check the buttonlocal
PointTwo highlighted areas capturedready
SendTranscript + screen contexttogether

Three small moves

Three moves.

01

Press

Choose the shortcut that fits.

02

Speak and circle

Mark the pieces that matter.

03

Paste

Send the complete context.

Try the experiment

Make the next prompt clearer.

Download the Apple Silicon build and try it in your next agent session.

Get BetterVoice