Live status and recent incidents for the AI stack most product teams build on, in one place. Refreshed every 5 minutes from each vendor’s official status API or feed. Subscribe to the aggregated RSS feed to get every new incident in one stream.
Rollup health and per-component status for every tracked vendor. Click through to the source status page for the official incident thread.
No components reported.
View official status pageActive incidents first, then resolved incidents from the last 30 days across every tracked vendor.
Needs investigation — Sun, 10 May 2026 22:05:35 GMT
Every 5 minutes. The page calls each vendor’s official status API or RSS feed, normalizes the response, and caches the result for 5 minutes so a high-traffic moment never overwhelms the upstream services. The Last Refreshed label in the hero shows the exact timestamp of the latest pull.
Anthropic, OpenAI, and Cursor all run public Statuspage v2 APIs at status.anthropic.com, status.openai.com, and status.cursor.com. Google AI Studio runs on Google Cloud, so we pull AI-related services from status.cloud.google.com. xAI runs on Instatus and blocks JSON endpoints, so we read its public RSS feed at status.x.ai/feed.xml. We do not scrape, alter, or interpret the data beyond normalizing severity labels into a shared scale.
Cursor sits between most engineering teams and the underlying AI providers, so a Cursor outage takes your editor down even when the model APIs are healthy. We track it on the same page so you can answer "is it me, my IDE, or the model" in one glance.
Operational means no known issues. Minor means at least one component has a small degradation. Major means a component is partially unavailable. Critical means a major outage is in progress. The labels mirror Statuspage’s own severity scheme so they match what each vendor publishes on their own status page.
Use the Subscribe via RSS button in the hero. The feed includes active incidents and any resolved incidents from the last 7 days across every tracked vendor, in one stream, so you can wire it into Slack, Feedly, an email automation, or any RSS reader.
When two or more vendors are simultaneously reporting non-operational status and the earliest active incident started in the last 6 hours, we surface a banner at the top of the page. It is a fast tell that the issue might be cross-vendor rather than specific to one product, which often points at a shared upstream like cloud infrastructure or DNS.
We track vendors and their components, not individual model endpoints. A provider-level outage in OpenAI’s API or Anthropic’s API affects every model behind it. If you need model-specific reliability, the Made By Agents reference model database is the better resource and we plan to add per-model uptime as a follow-up.
Embed support is on the roadmap and will reuse the existing embed pattern from the AI Hardware Calculator and ROI Calculator. For now, link to the page directly or subscribe to the RSS feed and surface the latest incidents in your own UI.
