STATUS
Generated 8 November 2026 from the gateway's own data. For the live state of any model, ask the API: it never starts anything.
SERVICE
| COMPONENT | WHAT TO CHECK |
|---|---|
| API gateway | GET https://api.verdantbloom.bar/healthz answers ok |
| Each model | GET https://api.verdantbloom.bar/v1/taps/<model>/eta: warm, starting, cold or full |
| Website and accounts | GET https://verdantbloom.bar/api/health |
Models run on GPUs rented by the minute. A model that has not been used for 10 minutes is stopped and starts again on the next request (typically 62 s, up to about 956 s). When GPU providers have no free cards, a request answers no_capacity and nothing is charged.
Incidents and questions: bloom@verdantbloom.bar.