Claude outage history

Claude investigated elevated errors affecting purchases, plan upgrades, and cancellations on claude.ai beginning at Nov 14, 2026 2:58 AM UTC. Core functionality remained available. Purchases returned to normal shortly afterward, and delayed plan changes and payments were applied; customers who had recently paid were told to allow up to an hour for their plans to update and not to pay again. Claude monitored recovery and marked the incident resolved by 4:23 AM UTC.

Elevated errors affected claude.ai, the Claude API, and Claude Code from around 1:13 PM UTC, with users also reporting failed connections to files, tools, and integrations. The disruption spread across Claude connectors such as Docs and Sheets, as well as Claude Security, Managed Agents, Meetings, and mobile apps. Reports came from users in multiple regions, indicating broad rather than localized impact.

Claude reported recovery on claude.ai and the API by 2:38 PM UTC, and said errors on those surfaces and Claude Code had returned to normal by 3:33 PM UTC. Some connector and Microsoft 365 issues, delayed or skipped scheduled tasks, and other feature disruptions persisted during the recovery. Claude marked the incident resolved by 9:53 PM UTC.

Users faced authentication failures and elevated errors across Claude.ai, Claude Code, and the Claude Console beginning around 10:04 AM UTC. Claude deployed a fix and reported recovery for Claude Code and the Console, but another period of degraded performance began around 1:13 PM UTC, affecting Claude.ai, the Claude API, Claude Code, and a wider set of features including connectors, Claude Security, and Managed Agents. Reports came from users across multiple regions, with many describing connection failures and nonworking connectors.

Claude users faced failures across several incidents beginning Nov 12, 2026 5:53 PM UTC. A Microsoft 365 service incident caused mail, calendar, search, and other connector requests to fail, blocked some new connections, and also affected Claude for Outlook. The Microsoft 365 connector was still failing for many requests the next day, with Claude reporting signs of recovery at 3:43 PM UTC.

Users attempting to use Claude Fable 5.1 encountered degraded performance and elevated errors on requests starting at 8:33 AM UTC. The impact affected the Claude Fable 5.1 ecosystem, including Claude API (api.anthropic.com), Claude Code, Claude Cowork, and claude.ai, resulting in slower responses and intermittent failures for affected users.

A fix was deployed and monitoring began at 8:43 AM UTC, with services returning to normal by 9:03 AM UTC, resolving the incident after about thirty minutes.

Users attempting to list or search Google Calendar events via Claude experienced elevated errors beginning 4:23 PM UTC. The disruption affected the Google Calendar connector across Claude.ai, Claude Code, and Claude Cowork, causing calendar-related requests to fail for some users and impacting workflows that rely on calendar integration.

Status updates indicated ongoing investigation with involvement from the upstream provider. By 7:48 PM UTC there were signs of recovery as error rates returned to normal, and all Claude components returned to full operation by 8:08 PM UTC. The incident spanned roughly four hours, during which users relying on calendar functionality likely experienced intermittent failures while listing or searching events.

An outage affected Claude's Gmail integration, causing elevated errors when users tried to search or read Gmail messages. The issue appeared to be isolated to the Gmail connector, with Claude.ai, Claude Code, and Claude Cowork remaining functional. The incident began at 4:03 AM UTC, and official updates documented ongoing investigation, including a follow-up at 4:08 AM UTC. By 5:53 AM UTC, the Gmail connector issues were resolved and Claude components returned to normal, ending a degraded period that lasted nearly two hours.

Claude Sonnet 5.5 began experiencing elevated errors at 3:13 PM UTC, impacting multiple Claude components including claude.ai, Claude API (api.anthropic.com), Claude Code, and Claude Cowork. Users attempting to use Sonnet 5.5 could have seen failed or slower responses while engineers investigated the issue. The degraded state persisted until 3:53 PM UTC when services were restored to normal operation.

An outage affected Claude Sonnet 5.5, causing elevated errors and some requests to fail or time out. The incident began at 9:18 AM UTC and impacted Claude Code, claude.ai, Claude API (api.anthropic.com), and Claude Cowork, all of which were listed in a degraded state before recovery. A user in Brampton, Ontario reported server not responding around 9:20 AM UTC.

Status updates on Claude's status page indicated engineers were investigating elevated errors on Claude Sonnet 5.5, with the affected components returning to normal by 9:38 AM UTC. The incident lasted about twenty minutes and affected users attempting to use Claude Sonnet 5.5 during that window, rather than the broader Claude platform.

A brief outage affected Claude Fable 5.1, causing elevated errors on requests to Claude Fable 5.1 and impacting multiple Claude services. The outage touched the Claude API (api.anthropic.com), Claude Code, Claude Cowork, and claude.ai. The incident began at Oct 25, 2026 8:18 AM UTC and was resolved by 8:33 AM UTC, spanning about 15 minutes.

Official updates noted that the cause had been identified and a fix was being implemented, with affected components returning to normal as the repair completed by the end of the window: Claude API (api.anthropic.com), Claude Code, Claude Cowork, and claude.ai moved back to the up state.