Status
Service health across the platform.
Availability is reported per component. When a provider degrades, automatic routing moves requests to the next healthy model in your chain.
1 component degraded
Provider Connections — Elevated latency on one upstream provider Requests routed through a fallback chain are unaffected.
Components
Current state, reported per service.
| Service | Status |
|---|---|
| API GatewayAll regions accepting requests | Operational |
| AuthenticationKey validation normal | Operational |
| RoutingFallbacks healthy | Operational |
| DashboardConsole responsive | Operational |
| Provider ConnectionsElevated latency on one upstream provider | Degraded |
- API GatewayOperationalAll regions accepting requests
- AuthenticationOperationalKey validation normal
- RoutingOperationalFallbacks healthy
- DashboardOperationalConsole responsive
- Provider ConnectionsDegradedElevated latency on one upstream provider
Uptime
Rolling availability per component.
| Component | 24 hours | 7 days | 30 days | 90 days |
|---|---|---|---|---|
| API Gateway | 100% | 100% | 99.99% | 99.98% |
| Authentication | 100% | 100% | 99.99% | 99.98% |
| Routing | 100% | 100% | 99.99% | 99.98% |
| Dashboard | 100% | 100% | 99.99% | 99.98% |
| Provider Connections | 99.94% | 99.91% | 99.88% | 99.90% |
Recent incidents
Post-incident summaries for the last 30 days.
- Elevated latency on one upstream providerSep 12, 2026 · Fallback routing moved affected requests automatically. No customer action was required.Resolved
- Console analytics delayed by 20 minutesAug 30, 2026 · Aggregation lagged behind request ingestion for roughly two hours. No requests were lost.Resolved
How to check for yourself
Two ways to verify the gateway is reachable from where you run.
- 1List models
A
GET /v1/modelscall returns the catalog with your key. A 401 means the key, not the service. - 2Watch your own logs
Request status and latency for your workspace are in the Activity console.