Jun 23, 2026, 02:19 PM → Jun 23, 2026, 04:44 PM · 2h 25m
This incident has been resolved.
We have seen error rates broadly return to normal success levels across all models. We saw elevated error rates on requests to multiple models from 7:08am PT / 14:08 UTC through 8:33am PT / 15:33 UTC.
We are continuing to monitor for any further issues.
A fix has been implemented and we are monitoring the results.
The issue has been identified and a fix is being implemented.
We are currently investigating this issue.