Is ElevenLabs Down?

Between 09:15 and 11:15 UTC on Nov 4, some Agent conversations in the EU using Eleven v4 Turbo had choppy audio, with parts of the agent's speech missing. This was caused by Eleven v4 Turbo generating audio slower than real time under high load. We have rolled out a fix that increases v4 Turbo throughput, and we are adding alerting to detect this earlier.

Some Agent conversations in the EU using Eleven v4 Turbo are experiencing choppy audio, with parts of the agent's speech missing. This is caused by Eleven v4 Turbo generating audio slower than real time under high load.

Between 10:10 and 11:05 UTC on Nov 3, some Agent conversations in the EU using Eleven v4 Turbo had choppy audio, with parts of the agent's speech missing. This was caused by Eleven v4 Turbo generating audio slower than real time under high load. The issue recurred on Nov 4, and we have since rolled out a fix that increases v4 Turbo throughput and are adding alerting to detect this earlier.

Some Agent conversations in the EU using Eleven v4 Turbo are experiencing choppy audio, with parts of the agent's speech missing. This is caused by Eleven v4 Turbo generating audio slower than real time under high load.

The issue affecting image and video generation, Flows, and MCP in ElevenCreative has been resolved. Generations are completing normally.

The elevated latency and timeouts affecting image and video generation, Flows, and MCP in ElevenCreative have improved. We are monitoring to confirm that generations are completing normally and will post an update once recovery is confirmed.

We are investigating elevated latency and timeouts affecting image and video generation, Flows, and MCP in ElevenCreative. Some generations may get stuck or fail to complete. Our team is actively looking into the cause and will post updates as we learn more.

Between 03:25 and 04:44 UTC on October 13, 2026, some requests to create an Instant Voice Clone failed with an error. This affected the Voice Lab in the app and the Voices API endpoint (POST /v1/voices/add), when background noise removal was turned on. Clones created without background noise removal weren’t affected.

Our engineers have confirmed full recovery of the Conversation Search functionality in the web UI and the API.

Currently, loading conversations in conversation history without filters is fully functional. Our engineers are still working to fully restore search filter functionality. Requests to the [Text search](https://elevenlabs.io/docs/eleven-agents/api-reference/conversations/messages/text-search "Text search") endpoint may continue to see failures as we continue to work to mitigate the issue.

Conversations are now populating in the conversation history again. Text search of conversations isn't working yet; we're actively fixing it and will post another update once it's resolved.

We are investigating an issue in the EU Residency where recent conversations are not showing up in history. The conversation history is not lost; there is just a lag in displaying it. Please note that this doesn't affect actual conversations; everything is working correctly. We are working on getting the history back up.

Our engineers have deployed a fix that has fully restored Claude MCP functionality.

We are currently working to release a targeted fix to restore functionality to the Claude MCP tool

We are investigating an issue where the MCP connection is not working. Please note that this does not affect ElevenAgents, where everything is working correctly.

The issue has been resolved, everything is back at normal.

We have released a fix, the error rate is back at zero. We are monitoring the situation.

We're investigating a latency increase on the EU Residency server for Speech-to-Text models, affecting a small subset of requests.

Inbound calls are currently operational.

From ~07:00 UTC, a small percentage (<1%) of inbound Twilio calls failed to connect due to a configuration error on our side. Affected callers were unable to connect to an agent; outbound calls and other traffic were unaffected. The change has been rolled back and calls have been connecting normally since ~11:00 UTC. No customer action needed.

We have confirmed full issue resolution.

We have identified the root cause: A misconfiguration in our billing system caused provisioning of subscriptions to be delayed. Errors remain at baseline. We are continuing our work to ensure the root cause is fully addressed to maintain future system integrity.

As of 14:55 UTC, we have observed error rates return to baseline, and are continuing to investigate the root cause.

We are currently investigating evidence of issues related to payment-processing. We will publish further details when the problem's scope and impact are clear.

Remediation measures have been fully rolled out, restoring failure rates to baseline. The root cause has been addressed, and the issue is now fully resolved.

Remediation measures have rolled out and instances of failures have returned to baseline. We are continuing to investigate further measures that may be necessary to completely mitigate the core issue.

Some Agent Platform conversations utilizing RAG are currently affected by RAG latency increases, in some cases resulting in slower RAG performance or intermittent call failures. Our team is currently working to resolve this.

We have identified the direct cause of the RAG failures, and are monitoring performance. At the moment, instances of call failures have returned to baseline. We are investigating further measures that may be necessary to completely mitigate the core issue.

Some Agent Platform conversations utilizing RAG are currently affected by RAG latency increases, in some cases resulting in slower RAG performance or intermittent call failures. Our team is currently working to resolve this.