Outbound calls dropping early across all Retell agents, Multiple Retell accounts: outbound calls hang up too soon

Support Help
1

I’m reaching out because I’m seeing a general issue across all my clients using Retell with Twilio via SIP trunking.

In summary:

  • I manage multiple Retell accounts and multiple Twilio projects, for different clients.

  • All of them use voice agents connected to Twilio through SIP trunking.

  • In all these environments, outbound calls are currently dropping after a short time (around a few tens of seconds), even though they used to work fine before and no configuration changes were made on my side.

  • This behavior is consistent across:

    • Different Retell accounts

    • Different Twilio accounts

    • Different agents and prompts

    • Calls started directly from the Retell dashboard and calls started via the Retell API (e.g. through n8n)

On the Twilio side:

  • Calls are shown as Completed, with normal duration (around 30–40 seconds), no error codes, and good network/latency metrics.

  • “Who Hung Up” is often shown as caller, even though neither I nor the end user manually hang up the call.

On the Retell side:

  • Sessions show a similar duration (around 30–40 seconds), but the call ends prematurely, often with outcomes like user hangup or Unsuccessful, and without the agent reaching its natural closing message.

  • The agent usually manages to play the initial greeting and the first question, then the call just drops without farewell, even though the agent settings allow much longer calls (e.g. End Call on Silence = 1 minute, Max Call Duration = 10 minutes).

Because:

  • This is happening across all my clients and agents,

  • Different Twilio accounts and SIP trunks are affected,

  • Twilio logs show no technical errors,

it looks like there may be a generic issue in the Retell ↔ Twilio SIP trunking integration, rather than a configuration problem specific to one agent or one account.

Could you please:

  1. Check if there is any ongoing incident or regression affecting outbound calls via SIP trunking?

  2. Review your internal logs for recent calls from my projects to see why Retell is ending them early, even though Twilio reports them as successful and with no errors?

  3. Let me know if there are any recommended changes or temporary workarounds for Retell + Twilio SIP trunking to keep calls from dropping so quickly.

I can provide multiple call_id (Retell) and Call SID (Twilio) examples from different accounts if needed.

Thank you for your help.

2

I understand you’re experiencing widespread call drops across multiple Retell and Retello accounts using Twilio SIP trunking. Based on the available sources, here are key troubleshooting steps:

Configuration Verification

Check Termination SIP URI(1)

  • Ensure your Termination SIP URI has no spaces(1)
  • Use a localized termination URI near your region(1)
  • Verify you’re using the correct format (typically ending with .pstn.twilio.com for Twilio)(1)

Authentication Setup(1)
Since Retell’s SIP server doesn’t have a static IP, you need to either:(1)

  • Whitelist all IP addresses in the range: 18.98.16.120/30 (All regions), 143.223.88.0/21 (certain US traffic), 161.115.160.0/19 (certain US traffic)(2)
  • Or use credential-based authentication with username and password(1)

Transport Method(2)
Verify your transport method is properly configured. Retell supports:(2)

For inbound calls with specific transport methods:(2)

  • TCP: sip:sip.retellai.com;transport=tcp(2)
  • UDP: sip:sip.retellai.com;transport=udp(2)
  • TLS: sip:sip.retellai.com;transport=tls(2)

Origination Setup(1)
Confirm your origination (inbound) is correctly configured with: sip:sip.retellai.com(1)

Agent Settings Review

Check your agent’s call duration settings:(3)

  • end_call_after_silence_ms: Minimum 10,000 ms (10 s), default 600,000 (10 min)(3)
  • max_call_duration_ms: Minimum 60,000 ms (1 min), maximum 7,200,000 (2 hours), default 3,600,000 (1 hour)(3)

Recommended Next Steps

  1. Check Twilio Logs: Review your Twilio elastic SIP trunk error logs for detailed error information(4)
  2. Verify Credentials: If you recently changed credentials, delete and re-import your numbers in Retell with updated authentication details(4)
  3. Review IP Whitelisting: Add Retell’s IP blocks to your Twilio IP Access Control List(4)

Since this is affecting multiple accounts systematically, I recommend:

  • Providing specific call_id examples from Retell
  • Checking Twilio’s elastic SIP trunk logs for patterns
  • Opening a support ticket with both Retell and Twilio with call examples

For additional support:

3

Thanks for the detailed response and the documentation you shared.

Just to clarify the situation:

  • These Retell + Twilio SIP trunking setups have been working perfectly for several weeks with the same configuration (termination SIP URI, authentication, transport, and agent settings).

  • The issue with calls dropping after a short time started recently, without any changes on my side to:

    • SIP termination/origination URIs

    • IP whitelisting or credentials

    • Transport method (we are still using the same one as before)

    • Agent settings (end_call_after_silence and max_call_duration are within the recommended ranges)

Regarding your suggestions:

  • Termination SIP URI: it has no spaces, uses the correct Twilio format for SIP trunking, and is localized to my region, and it has been unchanged since things were working.

  • Authentication: the same authentication method (IP/credentials) has been in place since the beginning; calls used to work fine with this setup.

  • Transport: the current transport method is the same as when everything was stable.

  • Agent settings: I’ve confirmed that end_call_after_silence_ms and max_call_duration_ms are well above the durations where calls are currently dropping.

On Twilio’s side:

  • The call logs show calls as Completed, with around 30–40 seconds duration, no error codes, and normal network metrics. There is no indication of a SIP or trunking error from Twilio’s side.

Given that:

  • The exact same configuration worked for weeks,

  • The problem started suddenly across multiple Retell and Twilio accounts,

  • And Twilio reports the calls as healthy,

it seems likely that something may have changed recently on the Retell side or in the Retell–Twilio SIP integration.

I’ll continue to gather more examples (Retell call_id and Twilio Call SID), but could you please check internally if there have been any recent changes or incidents that might explain why calls are now ending early, even though the previous configuration was stable?

Thanks again for your help.

4

Hello @harmandotvelasco

Can you share the call ids?

5

I’m an AI assistant knowledgeable about Retell AI. I don’t have context about your specific situation or which calls you’re referring to.

If you’re experiencing issues with your Retell AI calls, you can find call IDs in your Retell dashboard under the call logs section. Call IDs typically follow a format like call_6cf10655227b9f73619d31ed5dd or call_f831500a8e1540c03e0415e9fcd(1).

If you need help troubleshooting a specific issue, please share:

  • The call ID(s) you’re experiencing issues with
  • A description of the problem you’re encountering

This will help provide more targeted assistance with your Retell AI implementation.

6

Sure call_a173dc2d23fee20c401bbd25e27