Retell AI had 32 outages in the last 2 years totaling 6h 0m of downtime — averaging 1.3 incidents per month.
There were 32 Retell AI outages since June 19, 2025 totaling 6h 0m of downtime. Each is summarised below — incident details, duration, and resolution information.
19:42: there start to be reports on single / multi prompt agents having issues with function parameter being empty. 21:25: issue was identified and a hotfix is being rolled out. We are keep monitoring the issue.
This incident has been resolved.
Starting from around 6:30am, elevenlabs TTS is having a full scale outage: https://status.elevenlabs.io/incidents/01K0YAV1N4W7EZW1BMQCQJ4YJR Our platform traffic starts to fallback to other providers, and the volume of traffic started to break other TTS providers, and that's why you might have seen voice switching mid call, increased latency or even total silence. We're working with other TTS providers to bring in more capacity to support the fallback, and is monitoring closely on elevenlab's outage updates. Currently the error rate has returned back to normal. We will be actively following up on this.
This incident has been resolved.
Our telephony SIP provider confirmed there's an ongoing outage that led to phone call not connecting. We are escalating now, and is trying potential workarounds.
Telephony provider confirmed that this is related to an outage with Cloudflare DNS, which they use in certain internal components. A change is being applied to mitigate. related incident: https://www.cloudflarestatus.com/incidents/28r0vbbxsh8f
We are continuing to work on a fix for this issue.
The fix has been applied and phone calls are back to normal at the moment, we are keep monitoring.
We are continuing to monitor for any further issues.
This incident has been resolved.
3am: a culprit release was deployed, leading to some of the conversation flow agents stay silent after a function. It happens when a function and response is happening at same time, and is due to a message order issue before feeding to LLM. 8am: a fix is being deployed 8:30am: fix is fully deployed
Users are currently unable to sign up or create new workspaces on our platform. This issue stems from reaching the organization creation limit in Auth0, which is blocking new account provisioning. We are actively working with Auth0 to increase the limit
We are continuing to work on a fix for this issue.
The issue has been resolved, and new users can now sign up. Users who encountered sign-up issues today can proceed and log into the system without any problems.
9:30am: Incident started, where agents are not proactively responding to users when it has a low interruption setting. The behavior seen can include - Custom LLM not requesting responses - Retell agents not responding to user response 10:48am: A fix is being pushed, we start to see the issue get mitigated. 11:50am: Fix fully rolled out.
We have identified and fixed issues affecting transfer phone number with static number and missing call objects in custom tool invocations. Both issues have been fully resolved after 1 hour and 20 minutes of downtime.