Thanks — I hear you: outbound calls to one number are intermittently failing with "sip request timed out" very quickly (under a second), and you want more diagnostics or a way to increase the SIP timeout. Short answer • You cannot increase a user-configurable SIP request timeout in LiveKit Cloud. • The very-short timeout strongly indicates the remote SIP endpoint or trunk provider is rejecting or not answering the request rather than a slow-network timeout. What you can do next (steps) 1. Gather LiveKit-side signals ◦ Check webhook events (especially

participant_left

and its

reason

) for the failing calls to get the LiveKit close reason. ◦ Use Agent Observability / traces to correlate the session timeline, signaling spans, and any errors for each failed call. This will show where the failure occurs in the flow. 1. Check your SIP trunk/provider ◦ Ask your SIP trunk provider for logs for the failing call attempts (they often show SIP response codes or immediate rejects). ◦ Ask them to check for IP/domain blocking, ANI/caller-ID rejections, rate limits, or geographic routing issues to that destination number. 1. Test to isolate ◦ Try the same outbound number via a different trunk/provider (if available) to see if failures persist. ◦ If you can, test using a LiveKit Phone Number (removes third‑party trunk as a variable). 1. Share example events ◦ If you can share (paste) one or two webhook event payloads or trace IDs from Agent Observability for failed calls, I can look at the exact LiveKit reason and help interpret next steps. If you’d like, start by sending one webhook payload or a trace ID for a failed attempt and one that succeeded — I’ll compare them and suggest targeted next steps. Sources: Understand and Improve Agent Latency | LiveKit | Understand and Improve Agent Latency | LiveKit | Introducing LiveKit Phone Numbers: zero to ringing in 60 seconds | LiveKit | Deployment reliability on LiveKit Cloud | LiveKit | Agents telephony integration | LiveKit Documentation, with interactive elements