According to telemetry and internal testing, the temporary bump in resources was able to alleviate the issue. Additionally, we can confirm that the application load balancer improvement has been completed as well. While we continue to reduce resources back to operational levels and monitor the resources via telemetry, we will mark this incident as resolved.

During a planned reliability improvement to our application load balancers, a configuration issue for api.us.elevenlabs.io resulted in traffic not being properly distributed across all available zones. This led to backend overload and service degradation (primarily for website users).

To temporarily alleviate the issue, we have added additional resources to all the zones until the load balancer configuration can be safely remediated.