Text-to-Speech

Mona can speak her responses using text-to-speech (TTS). This works across all platforms — CLI, Telegram, Discord, and more.

TTS providers

Edge TTS (default)

Edge TTS is free and requires no API key. It uses Microsoft's online TTS service.

Setup

Already included in the base install. Configure the voice:

monoclaw config set tts.provider edge-tts
monoclaw config set tts.edge-tts.voice "zh-HK-HiuMaanNeural"

Available voices

List available voices:

monoclaw tts voices --provider edge-tts

Popular voices:

ElevenLabs (premium)

For the highest quality voices, including voice cloning.

Setup

cd ~/.monoclaw/monoclaw-runtime
uv pip install -e ".[tts-premium]"

monoclaw config set tts.provider elevenlabs
monoclaw config set ELEVENLABS_API_KEY "your-key"

See the dedicated ElevenLabs guide for detailed configuration.

Using TTS

CLI

/voice on

Mona speaks all responses. Press Ctrl+B to send voice messages.

Telegram

TTS is automatic for voice mode. To enable TTS for text replies:

monoclaw config set telegram.tts.enabled true

Discord

Enable in voice channels or for text replies:

monoclaw config set discord.tts.enabled true

TTS configuration

# ~/.monoclaw/config.yaml
tts:
  provider: edge-tts
  edge-tts:
    voice: "en-US-AriaNeural"
    rate: "+0%"      # Speaking rate
    pitch: "+0Hz"    # Pitch adjustment
  elevenlabs:
    voice_id: "21m00Tcm4TlvDq8ikWAM"
    model: "eleven_multilingual_v2"

Platform-specific settings

tts:
  cli:
    enabled: true
  telegram:
    enabled: false    # Only in voice mode
  discord:
    enabled: true
  slack:
    enabled: false    # Slack doesn't support voice well

Best practices

• Use Edge TTS for most cases — It's free and good enough
• Reserve ElevenLabs for presentations — Higher quality but costs per character
• Match voice to language — Use Chinese voices for Chinese text
• Adjust rate for long content — Slightly slower for technical explanations

Troubleshooting