Service disruption: Website and API

Maintenance Resolved

1. Resolved 2026-09-23 06:11 UTC

   The service was unavailable on 23 September from 05:57 to 06:07 UTC (about 9.5 minutes by our own checks). The cause was a routine update: it picked up a new release of a third-party component that was incompatible with our configuration, and the service failed to start after the restart. We pinned the known-good version of the component and deployed a fix; the service has been operating normally since 06:07 UTC. If a request failed or timed out during that window, please retry it. API keys, quotas and stored data were not affected. To prevent a repeat, we are pinning the versions of all third-party components.

2. Investigating 2026-09-23 05:58 UTC

   We are aware of an incident affecting this component and are investigating. Further updates will follow.