TL;DR

Real-time AI video chat that routes through your OpenClaw agent. Voice in, voice out, full agent context — like a video call with a knowledgeable assistant.

10-Second Pitch

• **What it does:** Adds live camera + voice to your OpenClaw agent using Groq Whisper (STT) and edge-tts (TTS)
• **Key win:** Your OpenClaw agent sees your camera and hears your voice in real time
• **Best for:** Hands-free AI assistance, accessibility use cases, ambient computing

How It Works

1. **STT** — Your voice goes to Groq Whisper for fast, accurate speech-to-text

2. **Agent** — The text is sent to your OpenClaw agent via chatCompletions API

3. **TTS** — The agent's response is synthesized with edge-tts (Microsoft's cloud TTS)

4. **Loop** — You hear the response and continue the conversation naturally

Setup

Install dependencies

pip install groq-whisper edge-tts openai

Configure OpenClaw agent endpoint

export OPENCLAW_ENDPOINT="http://localhost:18789"

export GROQ_API_KEY="your-groq-key"

Launch video chat

videochat-withme start --agent your-agent-name

What You Need

• OpenClaw agent running and accessible
• Groq API key (free tier available at console.groq.com)
• Microphone and camera

Pros / Cons

Verdict

`videochat-withme` is what ambient AI computing looks like. Your OpenClaw agent is already a powerful reasoning engine — now you can talk to it like a colleague. Worth installing if you want your agent to be genuinely accessible.