Status
- APIOperational · responded in 105ms
- ModelsNot enough traffic to say · no requests in the last 15 minutes
- DashboardOperational · you're reading it
Models
The model line above is measured from real requests in the last 15 minutes rather than from a test request of our own, so it reflects what customers are actually getting. An individual model can still fail on its own while the overall figure looks fine. If one is failing for you and this page looks clean, tell us — that is genuinely useful and we probably don't know yet.
What this page can't tell you
It runs on the dashboard, not the API, so it can report the API being down but it cannot report itself being down. If you can't load this page at all, that is its own answer.
The model line needs traffic before it will say anything. On a quiet morning it says it doesn't know, which is the honest answer — a handful of requests cannot tell the difference between an outage and somebody typing a bad model name.