Guides › How to Enable Streaming TTS in OpenCode
How to Enable Streaming TTS in OpenCode
Streaming TTS is enabled by default in the interrupt plugin. Responses are converted to speech sentence-by-sentence and played through your audio output.
1
Toggle with commands
Use /tts-on to enable TTS and /tts-off to disable it. TTS starts fresh for each session.
2
Change the voice
Set EDGE_TTS_VOICE env var to change the neural voice. Options include en-US-AriaNeural, en-GB-SoniaNeural, en-AU-NatashaNeural, and many more.
3
Adjust the speed
Set EDGE_TTS_RATE to control the speaking rate. "+25%" is the default (slightly faster than normal). "+0%" is natural speed. "+50%" is fast.
4
Speak specific text
Use /tts-speak followed by your text to make the plugin speak anything on demand. Useful for having the AI read code or error messages aloud.
Tips
• TTS plays through the default audio output device. Use system audio settings to route to headphones or speakers.
• When walkie-talkie recording is active, TTS automatically pauses to avoid feedback.
• edge-tts requires internet for voice data on first use per voice. Subsequent uses are cached.
Troubleshooting
If no audio plays, check your system audio output. On Linux, ensure pulseaudio/pipewire is running. On macOS, check the output device in System Settings.