Call bot lag time issues
Request for Native Gemini Multimodal Live Support in voice-call Plugin
Hey everyone, I’m running OpenClaw on bare-metal Ubuntu and using the voice-call plugin (Twilio + ngrok) to handle my assistant’s phone calls.
Currently, the plugin seems hard-coded to use the openai-realtime STT provider. While it works, the latency from the Twilio → ngrok → Local Machine → OpenAI round-trip is still quite high, making for a "walkie-talkie" experience.
I'm looking to switch to the Google Gemini Multimodal Live API to take advantage of its ultra-low latency and direct streaming capabilities. Does anyone have a roadmap for adding a google-live or gemini-live provider to the voice-call plugin? Or is there a way to bypass the OpenAI-specific streaming logic to pipe audio directly to Google’s live websocket?
Any help or pointers to a dev branch would be huge. Thanks!
Hey everyone, I’m running OpenClaw on bare-metal Ubuntu and using the voice-call plugin (Twilio + ngrok) to handle my assistant’s phone calls.
Currently, the plugin seems hard-coded to use the openai-realtime STT provider. While it works, the latency from the Twilio → ngrok → Local Machine → OpenAI round-trip is still quite high, making for a "walkie-talkie" experience.
I'm looking to switch to the Google Gemini Multimodal Live API to take advantage of its ultra-low latency and direct streaming capabilities. Does anyone have a roadmap for adding a google-live or gemini-live provider to the voice-call plugin? Or is there a way to bypass the OpenAI-specific streaming logic to pipe audio directly to Google’s live websocket?
Any help or pointers to a dev branch would be huge. Thanks!
The AI that actually does things. Emails, calendar, home automation - all from your favorite chat app. New shell, ready to help. The lobster way.
134,426Members
Resources
Similar Threads
Was this page helpful?
