| Date | Provider | Duration | Peak slowdown | Status |
|---|---|---|---|---|
| 2026-08-28 | OpenAI gpt-5.6-sol | 16 hours | — | Recovered |
| 2026-08-28 | OpenAI gpt-5-nano | 16 hours | — | Recovered |
| 2026-08-28 | OpenAI gpt-5.6-luna | 16 hours | — | Recovered |
Every incident here is measured, not reported to us by a provider. Duration is start to end (or start to now, for an open one); peak slowdown is the worst multiple of the 30-day median queue time we observed. The same history is available as JSON and as an Atom feed.