Next steps
For GPT-Live:
- Review delegation and tools for the backend prompt and application-owned context.
- Connect with WebRTC or WebSockets. See Telephony and SIP for phone integrations.
- Evaluate voice agents across conversation quality and verified task outcomes.
For Realtime:
- Review the earlier Realtime prompting guide for more
gpt-realtime-1.5examples. - Review the Realtime eval guide to test representative voice-agent behavior.
- Connect with WebRTC, WebSockets, or SIP.
- Learn the Realtime conversation lifecycle and review Realtime costs.