The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently...
Share of successful requests per carrier, measured over the last 5 minutes, 30 minutes and day. When one carrier dips, the router fails over to the next.
| Provider | Status | Last 5m | Last 30m | Last day |
|---|---|---|---|---|
| Operational | — | — | — |