Select an agent to start voice chatting
Talk to your AI agents in real time. Like WhatsApp calling, but for AI.
One command. Everything installs automatically. For non-tech users.
One container. Runs anywhere. For advanced users.
Clone the repo and run yourself. Full control.
This one command downloads, installs, and starts everything. Your mic becomes a live voice channel to your AI agents.
bash <(curl -fsSL https://loomy-voice-desktop.loomyloo.app/install.sh)
What this does:
~/.loomy-voice-desktop (hidden folder)After running, open this page in your browser and click "Connect." Your Mac's microphone becomes the input, your speakers the output. The proxy sits between you and Deepgram — it keeps your API key safe on your own machine, never in the browser.
Use this if you already have Docker installed.
docker run -p 98472:98472 -e DEEPGRAM_API_KEY=*** ghcr.io/loomisgitarrist/loomy-voice-desktop:latest
Or build from source:
git clone https://github.com/Loomisgitarrist/Loomy-Voice-Desktop.git && cd loomy-voice-desktop && docker build -t loomy-voice . && docker run -p 98472:98472 loomy-voice
For developers who want full control.
git clone https://github.com/Loomisgitarrist/Loomy-Voice-Desktop.git ~/.loomy-voice-desktop && cd ~/.loomy-voice-desktop && npm install
Then create your .env file:
cp .env.example .env && nano .env
Add your Deepgram API key, save (Ctrl+O, Enter, Ctrl+X), then start:
npm start
The proxy needs a Deepgram API Key to turn your voice into text and your agent's text into voice.
3bddee...).env file as DEEPGRAM_API_KEY=***Alternatives:
Once your proxy is running, come back here and click Connect below.
Select an agent to start voice chatting