Tavus Outage History
Tavus is up right nowPingoru has recorded 50 incidents for Tavus: 46 in the last 12 months and 11 in the last 90 days.
Show as a table
| Month | Incidents | Hours with an open incident |
|---|---|---|
| April 2026 | 7 | 16 of 720 (2%) |
| May 2026 | 1 | 0 of 744 (0%) |
| June 2026 | 2 | 2 of 720 (0%) |
| July 2026 | 5 | 18 of 744 (2%) |
| August 2026 | 6 | 5 of 744 (1%) |
| September 2026 | 3 | 5 of 720 (1%) |
| October 2026 | 0 | 0 of 588 (0%) |
Incidents, newest first
Minor March 25, 2026 Some conversations are encountering increased startup times 14m Conversational Video Interface
Started Mar 25, 2026, 01:24 PM UTC· resolved Mar 25, 2026, 01:39 PM UTC
- identified Mar 25, 2026, 01:24 PM UTC
The issue has been identified and a fix is being implemented.
- monitoring Mar 25, 2026, 01:37 PM UTC
We have rerouted traffic to healthy providers. Traffic has returned to normal.
- resolved Mar 25, 2026, 01:39 PM UTC
This incident has been resolved.
Minor March 10, 2026 Some conversations took longer time start –
Started Mar 10, 2026, 08:20 PM UTC· resolved Mar 10, 2026, 08:20 PM UTC
- resolved Mar 10, 2026, 08:20 PM UTC
We detected several conversations with start times exceeding 10 seconds. We have diverted traffic and are continuing to monitor the situation.
Notice March 6, 2026 Some conversations not starting up –
Started Mar 06, 2026, 01:27 PM UTC· resolved Mar 06, 2026, 01:27 PM UTC
- resolved Mar 06, 2026, 01:45 PM UTC
We have identified an issue with one of our GPU providers and shifted traffic away from them
Major February 17, 2026 List Replicas Endpoint 17m API
Started Feb 17, 2026, 02:06 AM UTC· resolved Feb 17, 2026, 02:23 AM UTC
- investigating Feb 17, 2026, 02:06 AM UTC
We are currently investigating an issue with our List Replicas endpoint
- resolved Feb 17, 2026, 02:23 AM UTC
API is back to operational
Minor February 12, 2026 Conversations slow to join 24m Conversational Video Interface
Started Feb 12, 2026, 07:45 PM UTC· resolved Feb 12, 2026, 08:10 PM UTC
- investigating Feb 12, 2026, 07:45 PM UTC
Some conversations might be slow to join. We're actively investigating
- monitoring Feb 12, 2026, 07:54 PM UTC
Traffic has been automatically shifted away from the faulty provider and service seems to be fully restored.
- resolved Feb 12, 2026, 08:16 PM UTC
This incident has been resolved.
Major January 23, 2026 Conversations are seeing delayed initiation 1h 41m Conversational Video Interface
Started Jan 23, 2026, 05:39 PM UTC· resolved Jan 23, 2026, 07:21 PM UTC
- investigating Jan 23, 2026, 05:39 PM UTC
We're seeing some conversations that are encountering delayed load times due to downstream GPU provider infrastructure outages. Our team is investigating and attempting to redirect traffic
- investigating Jan 23, 2026, 05:45 PM UTC
We are continuing to investigate this issue and are escalating the severity of this outage as we work towards a resolution
- identified Jan 23, 2026, 06:18 PM UTC
Three of our five downstream GPU providers began experiencing issues. Traffic is being actively rerouted away from affected providers, and we’re seeing performance improvements.
- identified Jan 23, 2026, 06:28 PM UTC
Traffic has been rerouted to healthy providers, and conversation load times are improving. Overall traffic is returning to normal levels.
- monitoring Jan 23, 2026, 06:44 PM UTC
Traffic has been successfully redirected to other providers and call initiation times have returned to normal. As traffic cleanup is underway, a sporadic load time may be seen in a small portion of conversations
- resolved Jan 23, 2026, 07:21 PM UTC
This incident has been resolved.
Minor January 21, 2026 Some conversations not initiating due to webRTC issue 3h 11m Conversational Video Interface
Started Jan 21, 2026, 07:27 PM UTC· resolved Jan 21, 2026, 10:39 PM UTC
- identified Jan 21, 2026, 07:28 PM UTC
A small portion of conversations are having issues initiating due to a WebRTC issue. We've been working closely with our webRTC partner to investigate the radius of impact and issue. The issue has been identified and a fix is being implemented.
- monitoring Jan 21, 2026, 10:09 PM UTC
The fix has been deployed with our WebRTC partner, and we’re continuing to closely monitor conversations.
- monitoring Jan 21, 2026, 10:38 PM UTC
We are continuing to monitor for any further issues.
- resolved Jan 21, 2026, 10:39 PM UTC
This incident has been resolved.
Minor January 7, 2026 Increased wait times for replica to join conversations 2h 39m Conversational Video Interface
Started Jan 07, 2026, 08:36 AM UTC· resolved Jan 07, 2026, 11:16 AM UTC
- investigating Jan 07, 2026, 08:36 AM UTC
We are investigating an issue where replicas take a long time to join conversations on one of our providers.
- investigating Jan 07, 2026, 08:36 AM UTC
We are continuing to investigate this issue.
- monitoring Jan 07, 2026, 08:52 AM UTC
We have diverted traffic away to healthy providers and are monitoring for any remaining increased joining times.
- resolved Jan 07, 2026, 11:16 AM UTC
This incident has been resolved.
Notice December 18, 2025 Temporary Video Generation Outage 3m Replica Video Service
Started Dec 18, 2025, 09:54 PM UTC· resolved Dec 18, 2025, 09:57 PM UTC
- investigating Dec 18, 2025, 09:54 PM UTC
We observed a temporary outage affecting video generation caused by an underlying infrastructure provider issue. Our team is actively monitoring recovery and working to ensure full service stability.
- resolved Dec 18, 2025, 09:57 PM UTC
This incident has been resolved.
Notice December 6, 2025 Elevated error rates with Llama-4 due to downstream provider errors 53m Conversational Video Interface
Started Dec 06, 2025, 07:36 PM UTC· resolved Dec 06, 2025, 08:30 PM UTC
- identified Dec 06, 2025, 07:36 PM UTC
The issue has been identified and a fix is being implemented.
- identified Dec 06, 2025, 07:39 PM UTC
We are moving traffic away from this LLM provider. Groq's status updates can be found here: https://groqstatus.com/incidents/01KBTGF16X8GQMQQB9Z5Z3Z1R4
- resolved Dec 06, 2025, 08:30 PM UTC
This incident has been resolved.
Notice December 2, 2025 Llama-4 Elevated Error Rates 1h 22m Conversational Video Interface
Started Dec 02, 2025, 05:55 PM UTC· resolved Dec 02, 2025, 07:18 PM UTC
- identified Dec 02, 2025, 05:55 PM UTC
Underlying Llama-4 provider is having elevated error rates causing some conversations to end early. The underlying issue can be monitored here: https://groqstatus.com/ We are working on diverting traffic away from the provider.
- monitoring Dec 02, 2025, 06:01 PM UTC
The underlying provider has implemented a fix and performance is improving
- monitoring Dec 02, 2025, 06:43 PM UTC
We are continuing to monitor for any further issues.
- resolved Dec 02, 2025, 07:18 PM UTC
This incident has been resolved.
Minor December 2, 2025 Replica Training and Video Generation Audio Issue 2h 37m Replica Training Service, Replica Video Service
Started Dec 02, 2025, 03:19 PM UTC· resolved Dec 02, 2025, 05:56 PM UTC
- investigating Dec 02, 2025, 03:19 PM UTC
We are currently investigating an audio generation issue for Replica Training and Video Generation. CVI is not affected.
- resolved Dec 02, 2025, 05:56 PM UTC
This incident has been resolved
Major November 20, 2025 Intermittent errors on Conversation Creation 16m Conversational Video Interface
Started Nov 20, 2025, 04:58 PM UTC· resolved Nov 20, 2025, 05:15 PM UTC
- investigating Nov 20, 2025, 04:58 PM UTC
We are currently investigating this issue.
- resolved Nov 20, 2025, 05:15 PM UTC
Our firewall identified anomalous traffic patterns and the systems are now stabilized.
Notice November 20, 2025 LiveKit Service Disruption 4h 17m Conversational Video Interface
Started Nov 20, 2025, 12:50 AM UTC· resolved Nov 20, 2025, 05:08 AM UTC
- investigating Nov 20, 2025, 12:50 AM UTC
LiveKit is experiencing an uptick in rooms that fail to subscribe to audio/video tracks with conversations disconnecting shortly after. Customers can fall back to the Tavus full pipeline which is unaffected.
- investigating Nov 20, 2025, 05:07 AM UTC
We are continuing to investigate this issue.
- resolved Nov 20, 2025, 05:08 AM UTC
This incident has been resolved.
Major November 18, 2025 Conversation creation failing 6h 18m Conversational Video Interface, API, Customer Portal
Started Nov 18, 2025, 01:07 PM UTC· resolved Nov 18, 2025, 07:25 PM UTC
- investigating Nov 18, 2025, 01:07 PM UTC
There's an outage with CloudFlare(https://www.cloudflarestatus.com/incidents/8gmgl950y3h7) affecting our GPU Providers. We are working on a solution
- investigating Nov 18, 2025, 01:27 PM UTC
Conversations created via API should be getting back to normal. We are still facing issues with our platform
- monitoring Nov 18, 2025, 04:28 PM UTC
Services are coming back online. We are still seeing longer than normal queue times as we continue to work around the persistent cloudflare outage (https://www.cloudflarestatus.com/incidents/8gmgl950y3h7).
- resolved Nov 18, 2025, 07:25 PM UTC
Services are back to normal operations. Let us know if you still see issues [email protected] .
Minor November 13, 2025 Upstream LLM Provider Disruption 4h Conversational Video Interface
Started Nov 13, 2025, 11:00 PM UTC· resolved Nov 14, 2025, 03:00 AM UTC
- identified Nov 14, 2025, 05:44 PM UTC
We observed a temporary increase in `llm_bad_request` and `llm_internal_server_error` responses stemming from an upstream LLM provider disruption.
- identified Nov 14, 2025, 05:46 PM UTC
The impact was limited, and service performance has since fully recovered.
- resolved Nov 14, 2025, 05:47 PM UTC
This incident has been resolved.
Major November 12, 2025 Experiencing intermittent log in issues 20m Customer Portal
Started Nov 12, 2025, 07:30 PM UTC· resolved Nov 12, 2025, 07:50 PM UTC
- identified Nov 12, 2025, 07:30 PM UTC
We have identified the root cause and are working on the fix.
- monitoring Nov 12, 2025, 07:47 PM UTC
We have deployed a fix and are monitoring
- resolved Nov 12, 2025, 07:50 PM UTC
This incident has been resolved.
Minor November 12, 2025 Increased utterance to utterance latency for tavus-llama 1h 30m Conversational Video Interface
Started Nov 12, 2025, 06:45 PM UTC· resolved Nov 12, 2025, 08:15 PM UTC
- investigating Nov 12, 2025, 06:45 PM UTC
We are seeing elevated utterance-to-utterance latency on the tavus-llama llm
- resolved Nov 12, 2025, 08:15 PM UTC
This incident has been resolved.
Notice November 11, 2025 Conversation Start Failures (GPU Provider Degradation) 3h 7m Conversational Video Interface, API
Started Nov 11, 2025, 09:58 PM UTC· resolved Nov 12, 2025, 01:05 AM UTC
- investigating Nov 11, 2025, 09:58 PM UTC
We’re currently investigating elevated error rates affecting the initialization of a small percentage of conversations with one of our LLM GPU providers.
- resolved Nov 12, 2025, 01:05 AM UTC
This incident has been resolved.
Minor October 20, 2025 Some Conversations Not Joined By Replica due to Ongoing AWS outage 13h 5m Conversational Video Interface, Replica Training Service
Started Oct 20, 2025, 01:15 PM UTC· resolved Oct 21, 2025, 02:20 AM UTC
- monitoring Oct 20, 2025, 01:15 PM UTC
AWS outage causing issues one of our infra providers
- monitoring Oct 20, 2025, 01:21 PM UTC
One of our infra providers is affected by the ongoing AWS outage. The previous message incorrectly said a fix had been implemented - we shifted traffic to a different provider.
- monitoring Oct 20, 2025, 07:22 PM UTC
We are continuing to monitor for any further issues.
- monitoring Oct 20, 2025, 07:23 PM UTC
CVI is now operational. We will continue to monitor the issue.
- resolved Oct 21, 2025, 02:20 AM UTC
This incident has been resolved.
Major October 20, 2025 AWS outage causing Conversation Creation Timeouts 2h 30m
Started Oct 20, 2025, 07:30 AM UTC· resolved Oct 20, 2025, 10:00 AM UTC
- resolved Oct 20, 2025, 09:52 AM UTC
Some conversation creation requests were impacted by the AWS us-east-1 incident and resulted in a 504 Gateway Timeout. This affected CVI. The incident lasted from 10:00 UTC and lasted until 10:20 UTC.
Major October 14, 2025 Cartesia outage resulting in text to speech failures 23h 16m Conversational Video Interface
Started Oct 14, 2025, 12:00 AM UTC· resolved Oct 14, 2025, 11:17 PM UTC
- identified Oct 14, 2025, 08:51 PM UTC
Cartesia, a text to speech voice provider for CVI, is encountering an outage resulting in some conversations "not replying". Elevenlabs voice users are unaffected We're in close touch with the Cartesia team, and the Cartesia status page is linked here: https://status.cartesia.ai/incidents/8bb0366tl7jp
- monitoring Oct 14, 2025, 09:45 PM UTC
An update from Cartesia team: Traffic is now being served with elevated latencies, as theyve implemented a patch on the infrastructure side. They're continuing to work on a resolution. We're closely monitoring the situation.
- resolved Oct 14, 2025, 11:17 PM UTC
This incident has been resolved.
Minor October 3, 2025 Delayed Video Generation Callbacks 2d 1h
Started Oct 03, 2025, 07:00 PM UTC· resolved Oct 05, 2025, 08:00 PM UTC
- resolved Oct 06, 2025, 12:17 AM UTC
Some video generation callbacks didn't properly deliver between October 3–5 as the callback jobs remained in an "in-progress" state and did not send as expected. The issue has been fully fixed now, and all affected callbacks have been retriggered.