System status
Platform status
Checked live against the platform health endpoint each time this page loads, and again every 60 seconds.
Degraded
Running health checks…
How incidents are handled
- Render failures retry across a three-model ladder before surfacing an error, and a failed generation is refunded automatically.
- Backend incidents are recorded as telemetry events and pushed to the on-call webhook the moment a health probe degrades.
- Exports are generated client-side, so downloads keep working during backend maintenance.