Antigravity Voice Mode
Talk to Antigravity's coding agent out loud, powered by Gemini Live, in the Antigravity app and CLI.
Details
- Voice mode continues your current conversation, so the agent keeps earlier context, including typed messages. You can interrupt it by talking and steer or stop a task while it runs.
- Permission requests are read aloud and you can approve or deny them by voice. Destructive actions, such as deleting files or force-pushing, must be approved by clicking Allow on screen, though you can still deny them by voice.
- In the app, click Start Voice mode with an empty message box. In the CLI, run /voice live or press F6. Requires Antigravity app or CLI version 1.5.0 or later.
- Nine voices, speed from 0.75x to 2x, and the agent is quiet by default unless you ask for periodic updates. Voice mode reaches only the conversation it was started in.
- The blog post does not state plan tiers, regions, or pricing.
Best for
Steering a coding agent hands-free while it reads code, edits files, and runs commands.
When to use it
When you want to brainstorm, guide a multi-step build, or approve actions without typing.
Who benefits
Developers who use Antigravity and want to pair-program by voice.
Join the newsletter to get ahead with AI.
I share strategies, new modules, free guides, and tools to get you better at using AI in your work.
Leading a team? See team programs →