Hi, I have a BYO SIP trunk credential (f7495eed-f03b-4eb7-b2f8-620418490a13) configured with gateway vapi-pstn.nickpilgui.voximplant.com:5060 (Voximplant Platform, US region, resolves to 3.147.160.14). Problem: Voximplant support confirmed that all SIP INVITEs from Vapi arrive at 185.164.149.100 (their RU SBC), NOT at 3.147.160.14 (their US SBC where our application is hosted). This causes 500 Internal Server Error on every outbound call. We also ran tcpdump on a separate Asterisk server (155.212.228.185) which we configured as gateway — zero packets received from Vapi IPs (44.229.228.186, 44.238.177.138) despite optionsPingEnabled: true. Evidence: tcpdump on 155.212.228.185 — zero packets from Vapi Voximplant support checked 3.147.160.14 — zero packets from Vapi Voximplant support checked 185.164.149.100 — traffic FROM Vapi confirmed All calls end with sipStatus: 500, phoneCallTransport: pstn Recent call IDs: 019e31bc-22c4-7001-9760-963b56e83b31 (2026-05-16 17:00 UTC) 019e3100-3f45-7001-9db2-5ad01eb736c5 (2026-05-16 13:36 UTC) 019e30f2-7af5-722d-beb1-c1aa547d865d (2026-05-16 13:29 UTC) Questions: Why is Vapi sending INVITEs to 185.164.149.100 instead of the configured gateway? Is there a known issue with DNS resolution of gateway domains in BYO SIP trunk? What exactly is in the Request-URI and To header of the INVITE Vapi sends for these calls? Phone number ID: 22478d1b-7347-463a-ad6b-3bd37732f523 Credential ID: f7495eed-f03b-4eb7-b2f8-620418490a13 Org ID: e9e3e7da-f82b-4b6c-a32d-57f1f748c905

Hi there, Thank you for your message. Our team is currently out of the office. We operate Monday through Friday, from 9:00 AM to 8:00 PM Pacific Standard Time (PST). We’ll get back to you as soon as possible during our normal business hours. If your message is urgent, please mark it accordingly or include “URGENT” in the subject line, and we’ll do our best to respond promptly. Warm regards, Vapi Customer Support Team

This looks more like a SIP routing/DNS resolution issue on the BYO trunk layer since Vapi is sending INVITEs to the RU SBC instead of your configured US gateway. The fact that both tcpdump and Voximplant confirmed zero traffic on 3.147.160.14 strongly points to incorrect gateway resolution or Request-URI handling internally. I can help you debug this directly and trace the SIP flow end-to-end. Have you tested forcing the SBC IP directly instead of the hostname in the credential config? @Nicolay_Pilgui

Yes, we tested forcing 3.147.160.14 directly as IP in the credential gateway (not hostname). Voximplant support confirmed they see zero traffic from Vapi IPs (44.229.228.186, 44.238.177.138) on 3.147.160.14. All traffic goes to 185.164.149.100 (Voximplant RU SBC) regardless of what we put in the gateway field. We also tested a separate Asterisk server (155.212.228.185) as gateway with optionsPingEnabled: true — zero packets received via tcpdump. It seems Vapi completely ignores the gateway config and routes through its own internal Voximplant carrier instead. Credential ID: f7495eed-f03b-4eb7-b2f8-620418490a13 Phone number ID: 22478d1b-7347-463a-ad6b-3bd37732f523 Recent call ID: 019e31bc-22c4-7001-9760-963b56e83b31 @Matt

At this point, it really doesn’t seem like the issue is on your side anymore. If direct IP routing and even a separate Asterisk server receive zero SIP traffic, then Vapi is likely bypassing the BYO gateway completely and routing through their own internal Voximplant layer instead. The main thing now is getting Vapi to verify how the credential + phone number are mapped internally, especially the outbound SIP routing logic and Request-URI generation. I’d genuinely be happy to help you work through this properly. If you want, we can discuss it more privately and dig deeper into the SIP traces and routing flow together. @Nicolay_Pilgui

Hi, We investigated the issue and confirmed that this is caused by DNS geo-routing behavior rather than a Vapi transport bug. Your configured hostname:

vapi-pstn.nickpilgui.voximplant.com

is resolving differently depending on geographic location. From Vapi’s infrastructure, the DNS resolution currently returns:

185.164.149.100

(Voximplant RU SBC) instead of:

3.147.160.1

(the US SBC where your application is hosted) Because the INVITEs are reaching the RU SBC, which does not recognize/configure your trunk, Voximplant responds with SIP

500 Internal Server Error

. Recommended fix: Use the explicit SBC IP directly instead of the geo-routed hostname. Example:

curl -X PATCH "https://api.vapi.ai/credential/f7495eed-f03b-4eb7-b2f8-620418490a13" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "gateways": [{
      "ip": "3.147.160.14",
      "port": 5060,
      "outboundEnabled": true
    }]
  }'

If the existing credential does not reprovision correctly after PATCH, we recommend deleting and recreating the credential using the explicit IP from the start.