What Microsoft, GitHub, Cloudflare, Ubiquiti, Cisco Meraki, Palo Alto Networks, GitLab, OpenAI, Anthropic, and Google post on their public status pages, read every 5 minutes while something is open. Each incident shows what else was happening and what changed in the same products just before.
We have identified an issue which resulted in some organizations to be incorrectly paused for having reached their spend limit, resulting in refused requests on the Claude API, Claude.ai, Claude Code, Claude Security, and Claude Cowork. This is now resolved.
Around it2 other incidents at the same time · 4 earlier incidents · 1 change in the 14 days before
This incident has been resolved. From 17:00 PT on October 6 / 00:00 UTC on October 7 to 18:40 PT on October 6 / 01:40 UTC on October 7, many requests to load usage data on the platform.claude.com Usage page and the Admin API usage report returned errors.
1 earlier update
7 Oct 01:20 UTCInvestigatingWe are investigating elevated errors loading usage data on the platform.claude.com Usage page and the Admin API usage report endpoint since 5 PM PT. Messages API requests are not affected. We will provide an update as soon as possible.
Around it1 other incident at the same time · 5 earlier incidents
The issue affecting Claude Opus 5.5 has been resolved.
2 earlier updates
6 Oct 12:37 UTCIdentifiedWe have identified the cause of elevated errors on requests to Claude Opus 5.5 and are working on a fix. We will provide an update as soon as possible.
6 Oct 12:24 UTCInvestigatingWe are investigating elevated errors on requests to Claude Opus 5.5. We will provide an update as soon as possible.
Around it1 other incident at the same time · 4 earlier incidents · 1 change in the 14 days before
The issue affecting Claude Mythos 5.1 and Claude Fable 5.1 has been resolved. Impact occurred from 05:40 PT / 12:40 UTC to 06:10 PT / 13:10 UTC.
1 earlier update
5 Oct 13:05 UTCInvestigatingWe are investigating elevated errors on requests to Claude Mythos 5.1 and Claude Fable 5.1. We will provide an update as soon as possible.
Around it3 other incidents at the same time · 3 earlier incidents · 1 change in the 14 days before
Bars compare how long each incident lasted, up to three days; a lighter bar is a lower bound and a hatched one is unknown. Older incidents are kept for 90 days on the timeline.
Azure post-incident reviews
What went wrong in major incidents and what Microsoft is changing