ElevenLabs Outage History

Starting at 03:50 UTC, a portion of SIP Trunks imported into ElevenLabs experienced failures when SIP calls were initiated. Mitigation measures were implemented at 16:45 UTC to resolve the issue. The service is now under continued monitoring to ensure full resolution.

ElevenLabs experienced voice generation failures affecting the "George" voice (ID: JBFqnCBsd6RMkjVDRZzb) starting at 19:30 UTC, with similar issues reported for legacy voices. The service provided a temporary workaround by directing users to use an alternative voice ID (6WwXjDDEMyNmFG95zycZ) from the library and recommending switching from legacy voices to non-legacy alternatives. The incident was resolved after 1.8 hours with the cause identified and a fix implemented.

ElevenLabs experienced an issue with their Conversational v3 Model (Alpha) where some calls became stuck in an "in progress" state, making conversation recordings, transcripts, and analytics unavailable for those calls. The incident lasted 4.5 hours, with the root cause identified and a fix deployed. Affected calls were automatically transitioned to "failed" status within 48 hours.

ElevenLabs experienced increased conversation latency affecting Gemini models for 1.3 hours, with conversations experiencing elevated response times but not failing due to fallback mechanisms. The issue was identified as originating from their cloud provider, and ElevenLabs worked with them to resolve it. Error rates decreased and the incident was resolved after cloud provider confirmation.

ElevenLabs experienced elevated 429 "system_busy" errors affecting their Speech-to-Text API for 7 minutes. Users encountered increased rate limiting when attempting to use the STT service. The incident was resolved after the brief 7-minute duration.

ElevenLabs experienced website slowness and API request latency for 5.2 hours due to an underlying cloud provider issue. The incident primarily affected the Agents Platform with degraded conversation initiation performance and caused intermittent website slowness, while TTS API, STT API, and Dubbing API continued working normally. The engineering team implemented changes to affected services that resulted in recovery and improved performance, with continued monitoring to confirm full stability.

ElevenLabs experienced an issue with their v3 conversational text-to-speech model on the Agents platform for 2.1 hours. The company investigated the problem and deployed a fix that restored normal error rates. The service is now operating normally with ongoing monitoring.