Set Up Voice Messages for Your Agent
Talk to your AI agent and hear it respond. Voice notes on Telegram, WhatsApp, or Discord — hands-free AI assistance.
How It Works
🎤
You Speak
Send a voice note on Telegram/WhatsApp
🧠
Agent Thinks
Voice → text (Whisper) → AI response
🔊
Agent Speaks
Response → voice (TTS) → audio sent back
Speech-to-Text (STT) Options
Groq Whisper ⭐ Recommended
Free tier available. Fastest transcription. Get a key at console.groq.com
OpenAI Whisper
$0.006/minute. Most accurate for noisy environments.
Text-to-Speech (TTS) Options
Edge TTS ⭐ Free
Microsoft's TTS engine. 100+ voices, multiple languages. Completely free. Default choice.
ElevenLabs
Most natural-sounding voices. 10k chars/mo free, then $5/mo+. Best for premium use cases.
Configuration
Add to your OpenClaw config:
{
"voice": {
"stt": { "provider": "groq" },
"tts": { "provider": "edge", "voice": "en-US-MichelleNeural" }
}
}Restart after config change: openclaw gateway restart