VoiceDOM Add to Chrome

Chrome extension

UI feedback that moves as fast as you speak.

Click any element, speak naturally, and get transcribed instructions with exact context for Cursor, Codex, Claude Code, and other AI coding agents.

BYOK · No subscription · Start free

Install in under a minute. Free tier uses Groq — bring your own API key.

17s demo · pick → record → transcribe → copy agent-ready markdown

Stop typing UI feedback. Point and speak instead.

VoiceDOM links your spoken request to the exact DOM element — so agents get clear text, selectors, and page context without back-and-forth.

Before

2–5 minutes rewriting visual feedback

Still unclear which element changed or what the agent should touch.

After VoiceDOM

20–40 seconds to speak and paste

Structured markdown with selector, URL, component name, and transcript — ready for your agent.

From idea to agent in under a minute

Three steps. No layout shift. Annotations saved per site.

  1. 1

    Select any element

    Crosshair picker with DevTools-style highlight. React component names when available.

  2. 2

    Speak your change

    Mic permission once. Transcribe with your Groq key on Free, or any provider on Pro.

  3. 3

    Ship to your agent

    Copy structured markdown or let Pro MCP read annotations live — no copy-paste.

Built for real web app work

Non-destructive picker

Highlight overlay with zero layout shift. Works on dynamic SPAs and modern frameworks.

Voice → text, instantly

Groq Whisper on Free with your own key. Pro unlocks OpenAI, Deepgram, ElevenLabs, and more.

Agent-ready export

Markdown with selector, XPath, tag, component name, URL, and optional screenshot refs.

Saved per site

Annotations persist in local storage, grouped by hostname. Review and clear from settings.

Component detection

Fiber introspection shows <LoginButton /> alongside the CSS selector.

MCP in Pro

Claude Code reads your annotations directly from a local MCP server — skip the clipboard. Setup guide →

Bring your own key

Your keys stay local. Pick any provider.

API keys live in your browser profile — never on our servers. Start with Groq on Free, then switch to OpenAI, Deepgram, ElevenLabs, or another supported provider whenever you want.

  • Keys stored locally
  • Switch anytime
  • No key on our servers

Start free. Upgrade once if you need more.

No monthly fee. Pro is a $19 lifetime unlock via ExtPay.

Free

$0

Forever — core workflow included

  • DOM picker + unlimited annotations
  • Groq transcription with your own API key
  • Markdown export to Cursor, Codex, Claude Code
  • MCP & screenshots (Pro)
Add to Chrome — free
Most popular for teams

Pro

$19 lifetime

One-time purchase — no recurring charges

  • Everything in Free
  • MCP server — agents read annotations live
  • BYOK for all transcription providers
  • Element screenshots in exports
Get Pro — $19 lifetime

Unlock inside the extension after install

Common questions

Is this only for voice notes? +

No. Voice is the input — the output is clean transcription linked to exact UI elements for fast AI handoff.

Which AI tools does it work with? +

Any tool that accepts text context: Cursor, Codex, Claude Code, and more. Paste markdown or use MCP in Pro.

Where are annotations and keys stored? +

Locally in your browser profile, grouped per hostname. Nothing leaves your machine unless you export it.

Does it work on dynamic web apps? +

Yes. Built for SPAs and modern frameworks with non-destructive highlighting and React fiber detection.

What if transcription fails? +

Retry immediately or switch provider in settings. Your saved annotations remain either way.

Do I need a subscription for Pro? +

No. Pro is a one-time $19 lifetime purchase. Free tier has no time limit.

Point. Speak. Transcribe. Ship.

Turn UI feedback into agent-ready instructions in seconds.