AI API outages on August 11, 2026
UTC day · measured every 5 minutes from 4 regions · method
On August 11, 2026 (UTC) our probes attributed 7 outage(s) to 1 provider(s): OpenAI API: 7.
This page is history. Is it down right now?
Live status, updated every 30 minutes: OpenAI API · all AI APIs, last 72 hours
Was the OpenAI API down on August 11, 2026?
Yes, measured. 7 outage(s) attributed to OpenAI: host probes failed from up to 4 of 4 regions at once while no other provider was failing, first at 18:25 UTC, at least 105 minutes observed in total. OpenAI’s status page lists nothing that day.
| UTC | What failed | Regions | Duration | Failed probes | Verdict | Error |
|---|---|---|---|---|---|---|
| 15:50–16:40 | host probes | 3 of 4: Europe (Germany), South America (São Paulo), US (Central) | ≥50 min | 12 | shared with 2 other provider(s) — not attributed | timeout |
| 17:55–18:00 | host probes | 3 of 4: Asia (Tokyo), Europe (Germany), US (Central) | ≥5 min | 3 | shared with 2 other provider(s) — not attributed | timeout |
| 18:25–18:40 | host probes | 2 of 4: Asia (Tokyo), South America (São Paulo) | ≥15 min | 3 | provider outage | timeout |
| 19:25–19:35 | host probes | 2 of 4: Asia (Tokyo), South America (São Paulo) | ≥10 min | 2 | provider outage | timeout |
| 20:10–20:35 | host probes | 4 of 4: Asia (Tokyo), Europe (Germany), South America (São Paulo), US (Central) | ≥25 min | 8 | provider outage | timeout |
| 21:20–21:20 | host probes | 2 of 4: Europe (Germany), US (Central) | <5 min | 2 | provider outage | timeout |
| 22:05–22:15 | host probes | 3 of 4: Asia (Tokyo), Europe (Germany), US (Central) | ≥10 min | 3 | provider outage | timeout |
| 22:30–22:55 | host probes | 3 of 4: Asia (Tokyo), South America (São Paulo), US (Central) | ≥25 min | 6 | provider outage | timeout |
| 23:25–23:45 | host probes | 4 of 4: Asia (Tokyo), Europe (Germany), South America (São Paulo), US (Central) | ≥20 min | 5 | provider outage | timeout |
13 single-region failure(s) were also recorded that day (yi-01ai 7, openai 2, qwen 1, friendli 1, stepfun 1, sambanova 1). A failure seen from one region cannot be separated from a problem on that network path, so none of them is called an outage.
How is an outage told apart from a network problem?
Each provider is probed every five minutes from 4 independent regions. A failure is attributed to the provider only when it is seen from two or more regions at once and no other provider fails in the same minutes; a failure shared with other providers has a cause above all of them and is not blamed on anyone. Durations are lower bounds: we see the first and last failed probe, not the real start and end. Status-page entries are the provider’s own words, shown next to the measurement, not instead of it.
Previous: Aug 9 · Last 72 hours · Next: Aug 12
Machine-readable: incidents.json · Licence: CC BY 4.0