On the morning of October 25, 2026, Claude had a bigger problem that lasted about 22 minutes, from 06:14 UTC to 06:36 UTC. Anthropic has since marked it resolved. No detailed cause has been published.
Who this kind of outage hits
If you build on Claude, your automations, your AI-assisted workflows, your app features that call the API, all of them go quiet at the same moment. You probably did not find out from a log or an alert. You found out because a customer wrote in to say something was broken, or because you happened to open the tool yourself and noticed nothing had run. That gap, between when the trouble started and when you heard about it, is where the damage happens.
Why it is especially rough if you are not an engineer
There is no error on your screen. There is no red light. The work just stops moving. A workflow that should have fired at 06:15 simply did not. You have no way to know whether it is your setup, your account, or the tool itself. So you start poking around, rerunning things, wondering if you broke something. By the time you figure out it was Claude’s side, your customer has already waited, or already complained.
That is the quiet part of a short outage. Twenty-two minutes sounds small. But if a customer-facing feature stopped working during those 22 minutes and nobody told you, the conversation you have later is harder than it needed to be.
Timeline
- 06:14 UTC - Claude starts having a bigger problem.
- 06:36 UTC - Claude marks the issue resolved.
- Total disruption - about 22 minutes.
How a watcher catches this before your users do
NoCrash reads Claude’s own public status page every minute. The moment that page flips from working to having trouble, NoCrash sends you a plain-language heads-up. Not a raw status code. Not a wall of technical text. A short note that says, in plain English, that Claude is having a problem right now.
That means instead of hearing it from a customer, you hear it within a minute of Anthropic’s own public report. You can pause a campaign, set an expectation, or just know to stop debugging your own setup. The signal arrives in one place alongside everything else you build on.
NoCrash also watches the things you ship. If you have n8n workflows, it watches those. If you have an app, you can give it a URL or drop in a small JS snippet and it will watch that too. So if something goes quiet on your own side, that surfaces as well. What it does not do is find the outage before Anthropic’s own status page does. It reads that page and tells you fast. That is the honest version of what it is.
For the authoritative account of this outage, go to Anthropic’s official status page: https://stspg.io/s3wd6gmdsdhn