Skip to content

Anthropic

Errors hit Claude Sonnet 5, Haiku 4.5 and Claude Code

Written by an AI model from the linked sources and checked against them by a second model. How we work

All about: Claude Code versions and changes →

Anthropic posted two notices at once on August 18, both stamped 16:20 UTC. One said it was investigating “degraded performance affecting multiple models.” The other said it was looking into elevated errors on requests to Claude Mythos 5, Claude Fable 5, Claude Opus 5, Claude Sonnet 5, Claude Haiku 4.5 “and other Claude models,” according to the Claude status page.

The incident covers claude.ai, the Claude API at api.anthropic.com, Claude Code and Claude Cowork. Coding agents, API integrations and chat users are all affected. Anthropic had not posted a cause with either notice.

This is the third degradation hitting several models since August 5, 13 days earlier. That first one lasted 7 hours 8 minutes, and outage trackers recorded a peak of 85 user reports per hour. The August 12 incident ran 4 hours 16 minutes but peaked at just 2 reports per hour, so a long incident didn’t always mean many users noticed. Both were eventually marked resolved.

The August 5 incident hit the same three top models that lead this list, Mythos 5, Fable 5 and Opus 5. That time, the status page said at 07:05 UTC that the cause had been identified and a fix was in progress. The August 18 notices name more models and still describe the incident as under investigation, with Anthropic promising an update “as soon as possible.”

No comments yet. Be the first.

Join the conversation

Sign in to leave a comment. It appears after moderation.