AI API outages on August 31, 2026
UTC day · measured every 5 minutes from 4 regions · method
On August 31, 2026 (UTC) no AI API outage could be attributed to a provider: nothing failed from two or more regions at once without other providers failing alongside. Providers’ own status pages opened 3 entries (Anthropic, baseten).
This page is history. Is it down right now?
Live status, updated every 30 minutes: Claude API (Anthropic), baseten API · all AI APIs, last 72 hours
Was the Claude API (Anthropic) down on August 31, 2026?
Not by our measurements. No failure of Anthropic was seen from two or more regions at once. Anthropic’s own status page opened 2 entries that day (self-reported, listed below).
Self-reported on Anthropic’s status page, not verified by our probes:
- minor Degraded performance on claude.ai · resolved Aug 31, 17:52 UTC
- minor Degraded performance on claude.ai and Claude Code · resolved Aug 31, 20:36 UTC
Is the Claude API (Anthropic) down right now? Live status →
Was the baseten API down on August 31, 2026?
Not by our measurements. No failure of baseten was seen from two or more regions at once. baseten’s own status page opened 1 entry that day (self-reported, listed below).
Self-reported on baseten’s status page, not verified by our probes:
- major Elevated inference 500s for models in a single US-based cluster · resolved Aug 31, 07:02 UTC
Is the baseten API down right now? Live status →
40 single-region failure(s) were also recorded that day (yi-01ai 21, google 5, sambanova 3, baichuan 3, writer 2, friendli 1, groq 1, iflytek 1, siliconflow 1, featherless 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.
← Aug 30 · Last 72 hours · Sep 1 →
Machine-readable: incidents.json · Licence: CC BY 4.0