Claude Outage History

Claude experienced a 21-hour outage affecting usage reporting across the Console, admin API endpoints, and analytics dashboards due to an internal data service issue. The outage caused errors and missing data in usage reports from February 26 ~02:00 to February 27 06:20 UTC, impacting claude.ai, console.anthropic.com, and api.anthropic.com. All endpoints have been restored with historical data now displaying correctly.

Claude Opus 4.6 experienced elevated errors for 24 minutes, affecting claude.ai, console.anthropic.com, and api.anthropic.com. A fix was implemented and the incident was resolved after monitoring confirmed the solution was effective.

The Claude Desktop app failed to start for some Windows users and macOS users on older releases due to an identified technical issue. The problem affected the claude.ai service and prevented affected users from launching the desktop application for 7.6 hours. The issue was resolved by releasing version 1.1.4328 with a fix, requiring users to download the latest version from the official website.

Claude Sonnet 4.6 experienced elevated errors for 1.1 hours, affecting claude.ai and console.anthropic.com. Users reported receiving "taking longer than usual" messages that would cycle through multiple retry attempts without ever providing responses. The service was restored to baseline error levels as of 5:10 PT.

Claude Sonnet 4.6 experienced elevated error rates affecting both claude.ai and api.anthropic.com for 16 minutes. The incident impacted users' ability to access Claude services normally. The errors were resolved and returned to baseline levels as of 23:27 PT / 7:27 UTC.

Claude Haiku 4.5 experienced elevated errors for 1.9 hours, affecting claude.ai, console.anthropic.com, and api.anthropic.com. Users encountered service disruptions with messages indicating Claude was temporarily unavailable. The issue was identified and resolved with a implemented fix.

Claude Sonnet 4.5 experienced elevated error rates affecting both claude.ai and api.anthropic.com for 16 minutes. The issue was identified and a fix was implemented, with error rates returning to baseline levels. Both the web interface and API services were impacted during this minor incident.

Claude Opus 4.6 experienced elevated error rates for 2 hours, affecting claude.ai, console.anthropic.com, and api.anthropic.com. Users reported issues accessing the claude.ai website during coursework. The issue was identified and fixed, with error rates returning to normal levels.