Voice Communication
This skill enables voice communication on Telegram:
- Receive: Transcribe voice messages using faster-whisper
- Reply: Send voice replies using TTS
Incoming Voice (Automatic)
When receiving voice messages (.ogg files), use faster-whisper to transcribe:
from faster_whisper import WhisperModel
model = WhisperModel('base', device='cpu', compute_type='int8')
segments, info = model.transcribe('<file_path>', language='zh')
result = ''.join([s.text for s in segments])
Outgoing Voice (TTS)
Use the tts tool to send voice replies:
{
"action": "send",
"channel": "telegram",
"message": "<text>",
"asVoice": true
}
Or use the tts tool directly:
{
"channel": "telegram",
"text": "<text to speak>"
}
Language
- Input: Auto-detect or specify language (zh for Chinese)
- Output: Match user's language preference
Requirements
- faster-whisper:
pip install faster-whisper - TTS already configured in OpenClaw