Live  /  Claude  /  Incident

Claude: Elevated errors for Claude Opus 5.5

Claude · claude.aiMajor incidentResolved2026-10-06 12:43 UTC
Public — what happened

Summary

Claude reported a major incident in which requests to its Claude Opus 5.5 model were failing at a higher rate than normal, beginning at 12:24 UTC on 2026-10-06. By 12:37 UTC, Claude said it had pinpointed the source of the errors and was working on a fix, though it did not describe what that source was. The incident was marked resolved at 12:43 the same day.

Affected

claude.ai, Claude API (api.anthropic.com), Claude Code, Claude Cowork

As listed on Claude’s status page.

Timeline

Source of record: provider status page. Summary written by SquidGrid from the provider’s published updates; times in UTC.

Independent path health

The status page is the symptom. SquidGrid measures the paths that reach Claude against their own baselines — whether the routes carrying your traffic degraded, at which handoff, and who owns that stretch.

🔒 Sign in to see
Customers get the live Provider Assessment Report for Claude: its full incident track record, what tends to break, and recent public reports of problems.

Included with any paid plan.

Open the Claude report →

More from Claude

Covered in The Squid’s Ink, Week 42, 2026, our weekly internet health report.

The full archive is on the Claude status & history page.

Related: Claude status & history. SquidGrid is not affiliated with Claude.