AI API outages on October 22, 2026
UTC day · measured every 5 minutes from 4 regions · method
On October 22, 2026 (UTC) our probes attributed 3 outage(s) to 1 provider(s): Gemini API (Google): 3. Providers’ own status pages opened 4 entries (OpenAI, Anthropic).
This page is history. Is it down right now?
Live status, updated every 30 minutes: Gemini API (Google), OpenAI API, Claude API (Anthropic), writer API · all AI APIs, last 72 hours
Was the Gemini API (Google) down on October 22, 2026?
Yes, measured. 3 outage(s) attributed to Google: requests to gemini-flash-lite-latest failed from up to 3 of 4 regions at once while no other provider was failing, first at 12:01 UTC, at least 16 minutes observed in total. At least once Google’s own host answered with an HTTP 5xx.
| UTC | What failed | Regions | Duration | Failed probes | Verdict | Error |
|---|---|---|---|---|---|---|
| 06:41–07:01 | requests to gemini-flash-lite-latest | 3 of 4: Asia (Tokyo), Europe (Germany), US (Central) | ≥20 min | 5 | shared with 1 other provider(s) — not attributed | timeout |
| 12:01–12:01 | requests to gemini-flash-lite-latest | 2 of 4: Asia (Tokyo), Europe (Germany) | <5 min | 2 | provider outage | timeout |
| 12:35–14:06 | requests to gemini-flash-lite-latest | 4 of 4: Asia (Tokyo), Europe (Germany), South America (São Paulo), US (Central) | ≥90 min | 18 | shared with 2 other provider(s) — not attributed | http 503; timeout |
| 14:41–14:46 | requests to gemini-flash-lite-latest | 3 of 4: Asia (Tokyo), Europe (Germany), South America (São Paulo) | ≥5 min | 3 | provider outage | http 503; timeout |
| 15:21–17:41 | requests to gemini-flash-lite-latest | 4 of 4: Asia (Tokyo), Europe (Germany), South America (São Paulo), US (Central) | ≥140 min | 26 | shared with 5 other provider(s) — not attributed | http 503; timeout |
| 18:55–19:06 | requests to gemini-flash-lite-latest | 3 of 4: Asia (Tokyo), South America (São Paulo), US (Central) | ≥10 min | 3 | provider outage | timeout |
| 19:55–20:21 | requests to gemini-flash-lite-latest | 3 of 4: Asia (Tokyo), South America (São Paulo), US (Central) | ≥26 min | 4 | shared with 1 other provider(s) — not attributed | timeout |
Current Gemini API (Google) status
Was the OpenAI API down on October 22, 2026?
Not by our measurements. No failure of OpenAI was seen from two or more regions at once. OpenAI’s own status page opened 3 entries that day (self-reported, listed below).
Self-reported on OpenAI’s status page, not verified by our probes:
- minor Elevated ChatGPT conversation and Responses API errors · resolved Oct 22, 08:50 UTC
- minor Degraded Performance of Dots on Enterprise · resolved Oct 22, 20:55 UTC
- minor New Ad Accounts failing to provision admins · resolved Oct 23, 00:07 UTC
Was the Claude API (Anthropic) down on October 22, 2026?
Not by our measurements. No failure of Anthropic was seen from two or more regions at once. Anthropic’s own status page opened 1 entry that day (self-reported, listed below).
Self-reported on Anthropic’s status page, not verified by our probes:
- minor Elevated errors affecting Connectors · resolved Oct 22, 08:07 UTC
Current Claude API (Anthropic) status
Was the writer API down on October 22, 2026?
Not attributably. 1 multi-region failure(s) of writer coincided with failures of other providers, so the cause sat above writer and is not blamed on it.
| UTC | What failed | Regions | Duration | Failed probes | Verdict | Error |
|---|---|---|---|---|---|---|
| 15:40–15:41 | host probes | 3 of 4: Europe (Germany), South America (São Paulo), US (Central) | <5 min | 3 | shared with 2 other provider(s) — not attributed | timeout |
53 single-region failure(s) were also recorded that day (yi-01ai 23, featherless 8, google 7, doubao 3, iflytek 3, cerebras 2, baichuan 2, friendli 1, deepinfra 1, typesafe 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: Oct 21 · Last 72 hours
Machine-readable: incidents.json · Licence: CC BY 4.0