Investing elevated failure rate

Identified·Degraded performance

We identified an issue with our GPU cloud subprocessor. We have engaged our emergency fallback inference system and are seeing initial recovery albeit with degraded performance.

Wed, Jun 3, 2026, 01:32 PM

(1 hour ago)

Affected components

01:12 PM

Now

Updates

Identified

We identified an issue with our GPU cloud subprocessor. We have engaged our emergency fallback inference system and are seeing initial recovery albeit with degraded performance.

Wed, Jun 3, 2026, 01:32 PM

Investigating

We’re seeing an increase in failures and latency. Investigating now.

Wed, Jun 3, 2026, 01:12 PM(20 minutes earlier)