Hume AI

The issue has been resolved.

TTS is back up. An underlying deployment issue prevented this service from starting correctly. The issue has been resolved.

TTS is currently down. We have identified the issue and working on a fix.

Issue was resolved.

Seeing chats recover. Monitoring over the next few hours

We're seeing increase error rates on EVI. We're investigating.

Issue seems to be resolved. We mitigated the issue by increasing capacity and adjusting timeout/retry handling to improve reliability under load. Service health was confirmed once automated checks and live sessions were consistently succeeding.

EVI appears to be down in a majority of cases, and TTS also appears impacted.

We worked with the third-party provider to resolve the integration issue, restoring API availability.

A third-party integration experienced issues, leading to failures and timeouts in our APIs.

Integration issue has been fixed.

We identified an issue with a third-party integration and are actively working with them to remedy the issue.

A restart of the component mitigated the issue.

An issue with one of our models is causing some issues to EVI and TTS

Issue was resolved.

Long wait times on expression measurement due to load.

False Alarm: no downtime

We're seeing increased error rates on Voice Cloning

EVI and Octave are operational.

An outage with a downstream provider is impacting Octave and EVI.

High load on the service caused high queue times and failures for both EVI and Octave models. This has been resolved and the team is monitoring the fix.

We're actively investigating an outage affecting both EVI and Octave

The team identified the issue related to audio decoding from the Octave model and has deployed a fix.

We are investigating transient failures on the TTS service

This has been resolved and Octave is operational.

Issue resolved.

Actively monitored

TTS API currently impacted due to the AWS outage: https://health.aws.amazon.com/health/status We are working on bringing services back up.

Issue was resolved

Application website was down as some services were impacted by a global AWS outage.

Issue was resolved.

Discovered an issue related to our payments service for users that have organizations enabled.

Latencies have returned to their normal levels and APIs operating normally.

EVI and Octave services were timing out connecting to models from approximately 11 - 11:35pm EST. Issue seems resolved now, and team is monitoring

Issue has been resolved.

Monitoring the latency issue.

Issue has been identified and currently being worked on.

Timeout issue in model inference causing degraded performance.

Internal issue was fixed.

Internal issue found which caused requests to TTS to fail.

Services back up with cloudflare outage resolution.

APIs down due to Cloudflare outage.

Cloudflare back up and Hume services are back up

Most if not all Hume services are down currently due to a Cloudflare outage.

Issue was resolved

Service outage for web applications identified

Issue was resolved

Degraded performance for TTS-related products

Issue was resolved.

Due to an upstream inference provider we're seeing some failed requests on EVI and TTS.

An API key had expired and has been rotated. EVI appears functional again.

Our Speech-To-Speech API is currently experiencing timeouts, resulting in failed calls. Other services appear to be unaffected.

Multilingual transcription has been re-enabled. Service fully restored.

Multilingual transcription for EVI is currently down. Root cause is being investigated.

Everything is operating normally again. Monitoring will continue throughout the day.

A fix has been deployed. We're monitoring to make sure it's working as intended.

Certain voices are requiring data to be backfilled, which is causing excessive load on the system. A fix is currently in the works.

Issue was resolved

Service has fully recovered. The engineering team is continuing to monitor.

Seeing high error rates on TTS and EVI, seemingly due to an upstream inference provider. We are actively investigating.