AI Provider Status — Right Now
Real-time health check for every major AI API. Updated every 60 seconds. Bookmark this page so you know instantly when Claude, OpenAI, Gemini, or any of 29 other providers goes down.
Status by Category
Grouped by provider category. Click any provider for its full error library and hub page.
Recent Incidents
Latest status changes across all tracked providers. For full historical incident data, see individual provider pages.
Our Monitoring Methodology
We probe each provider's public API endpoints every 60 seconds using minimal-cost health-check requests: a simple ping completion for chat models, a status endpoint call for infrastructure providers, and a small generation task for image/audio APIs. Status classifications are based on error rate, response latency, and any reported provider incidents from official status pages.
Operational means normal error rate (<1%) and normal latency. Degraded means elevated errors (1–5%) or elevated latency (2–5x baseline). Slow means normal error rate but significantly elevated latency (>5x baseline). Outage means >5% error rate or complete endpoint unavailability. We correlate with each provider's official status page and flag discrepancies.
Status data on this page is display-only and should not be used as the sole basis for production alerting. For automated monitoring, integrate directly with your provider's official status API. For historical uptime, click any provider card above.