Voice

Core Voice

Capture voice from the microphone, review it in nvim, and launch an AI agent. Perfect for hands-free coding and quick prompts.

core voice opencode

View CORE-TERMUX Docs ←

Requirements

Termux:API

pkg install termux-api

Neovim

core install editor

Termux:API App

Download from DevCoreX

Commands

core voice

Show help

core voice <agent>

Capture → nvim → launch agent

core voice text

Capture → nvim → print to stdout

core voice !

Alias for 'text'

Supported Agents

Voice works with all major AI coding agents:

opencode

opencode run "prompt"

claude-code

claude -p "prompt"

codex

codex "prompt"

gemini-cli

gemini -p "prompt"

hermes-agent

hermes chat -q "prompt"

kimi-code

kimi -p "prompt"

mimocode

mimo run "prompt"

mistral-vibe

vibe --prompt "prompt"

openclaude

openclaude --bg "prompt"

pi

pi -p "prompt"

qwen-code

qwen -p "prompt"

text

Print prompt to stdout

Example Session

$ core voice opencode

    ➜ Listening through the microphone...
    ➜ Review the prompt in nvim, fix mistakes, then save and quit
    ➜ Launching opencode with prompt…

    # opencode opens with the voice-transcribed prompt