2 min read AI-generated

Claude Goes Down for the Tenth Time in Twelve Days — Anthropic's Infrastructure Hits Its Limits

Copy article as Markdown

Ten outages since June 5. Anthropic's servers can't keep up with demand — and this from a company whose revenue has quadrupled since year-end.

Featured image for "Claude Goes Down for the Tenth Time in Twelve Days — Anthropic's Infrastructure Hits Its Limits"

On Monday, Claude went down again. For the tenth time since June 5. Opus 4.8 and Haiku 4.5 threw errors, and a fix attempt at 2:00 PM ET initially failed. Anthropic confirmed on its status page that infrastructure is under “unprecedented strain.”

What’s behind the outages

The numbers explain the problem: Anthropic’s annualized revenue jumped from $9 billion at the end of 2025 to over $30 billion by April 2026. The number of enterprise clients spending more than $1 million annually doubled from 500 to 1,000 in under two months.

That’s growth that almost no infrastructure can keep pace with. Fable 5, Managed Agents, the Apple integration — everything is driving demand higher.

No post-mortems

What stands out: Anthropic hasn’t published a single public post-mortem for the June outages. No detailed root cause analysis, no engineering reports. Just rolling status page updates. For a company that emphasizes transparency as a core value, that’s notable.

What this means for users

If you use Claude in production — whether through the API, Claude Code, or Cowork — you need to expect instability. Ten outages in twelve days isn’t a one-off incident, it’s a pattern.

The silver lining: the outages are usually short and rarely affect all models simultaneously. Sonnet 4.6 and Fable 5 were often still reachable when Opus and Haiku had issues.

My take

Claude has become infrastructure — and that’s precisely the problem. When millions of developers, thousands of companies, and increasingly automated agents depend on a single service, “mostly available” isn’t good enough anymore.

Anthropic needs to deliver here. Not just by adding more servers, but by communicating more transparently. A company pulling in $30 billion in revenue can’t afford black-box outages — at least not if it wants to keep the trust it’s built within the community.

Sources: TechTimes, Thoughtworks