- name
- kai-realtime-voice
- description
- Real-time voice streaming via MiniMax WebSocket API. Use for low-latency voice conversations and streaming audio generation.
- metadata
- openclaw
- requires
- env
Kai Realtime Voice
Real-time voice streaming using MiniMax WebSocket API.
Setup
API key already configured in OpenClaw config. Uses same key as kai-minimax-tts.
Usage
Test WebSocket connection:
bash {baseDir}/scripts/kai_realtime.sh --testStream text to audio:
bash {baseDir}/scripts/kai_realtime.sh --stream "Hello world"Requirements
- Python3 with websockets library
- WebSocket support
- MiniMax API key (auto-injected)
Output
Audio streamed to stdout or saved to file.