Claude outage history

Claude Fable 5 experienced a brief outage with elevated errors when calling Claude.ai. The issue began at 11:58 PM UTC and was resolved by 12:28 AM UTC, lasting a little over 30 minutes. Claude's status page indicated that the cause had been identified and a fix was being implemented, and the Claude.ai component returned to normal as recovery progressed.

StatusGator detected issues across Claude services at Aug 14, 2026 8:20 PM UTC, roughly 11 minutes before Claude's status page acknowledged the disruption at 8:31 PM UTC. The outage affected Claude API (api.anthropic.com), Claude Code, and Claude Cowork, with user reports describing unresponsiveness and degraded functionality across tools, code environments, and cowork sessions. The incident began around 8:31 PM UTC and prompted multiple updates as the team investigated.

An initial fix was deployed by 8:48 PM UTC, and recovery progressed in fits and starts as individual components came back online. By 10:53 PM UTC all Claude services were restored to normal. The outage affected users across regions including the United States, United Kingdom, India, Denmark, Italy, and other locales, with reports of API requests failing, Claude Cowork sessions freezing, and container or document generation failures. The disruption lasted roughly two hours, with full restoration by the end of the window.

Claude experienced a brief partial outage of its status page when an invalid TLS certificate blocked access to status.claude.com. The disruption began at 10:43 AM UTC and was resolved by 10:58 AM UTC as engineers worked to fix the certificate issue. Claude provided an update acknowledging the problem and their ongoing efforts to restore normal access.

Users began reporting issues around 6:28 AM UTC as Claude became unavailable due to server overload. The outage produced widespread API failures with many requests returning 529 Overloaded errors or timing out, hindering access to Claude's API and related features. Activity spiked through the morning, peaking around 6:45 AM UTC and remaining elevated through 7:00 AM UTC as the impact spread across Claude's surfaces. StatusGator detected the outage ahead of Claude's acknowledgement. The incident persisted through the morning and was resolved around 8:51 AM UTC. Claude did not publish status updates in the provided feed during the incident, and users experienced degraded or unavailable access across API endpoints including Claude opus 5 and claude-code surfaces.

A brief disruption affected Claude Code on the Web and related services, with Claude Cowork reporting degraded performance for Claude Code on the Web sessions and connected products such as Cowork Remote, Routines, and Claude Code Mobile. Users likely saw slower sessions and intermittent issues while using Claude Code on the Web. The issue began at 12:38 AM UTC and was resolved by 12:53 AM UTC.

The service communication indicated an investigation followed by recovery, with Claude Cowork moving from a degraded state back to normal as the status changed to up. This was a brief incident and involved a small subset of users, based on observed reports during the window.

Users of Claude Mythos 5, Claude Fable 5, and Claude Sonnet 5 faced an outage beginning around Aug 13, 2026 2:38 PM UTC, with elevated errors across several Claude components. Reports described sites not loading, chat features failing, and API requests returning 529 overload errors. The disruption affected key Claude services including claude.ai, the Claude API at api.anthropic.com, Claude Code, and Claude Cowork, limiting access to the assistant and related tooling. The incident persisted for several hours, impacting user workflows and productivity as sessions and requests returned errors.

Starting at 1:53 PM UTC, Claude experienced degraded performance across multiple models and endpoints. The incident affected Claude API (api.anthropic.com), Claude Code, Claude Cowork, and claude.ai, causing elevated errors and failed requests for many users. Some users reported OAuth token verification failures and prompts to switch models, with the impact most noticeable on Claude Fable 5 as the afternoon progressed.

Status page updates showed ongoing remediation, with notes at 4:13 PM UTC and 5:33 PM UTC indicating errors across multiple models and continued efforts to restore normal operation. By 6:08 PM UTC, all components had returned to normal and the incident was considered resolved. User reports described authentication/token verification failures and model-selection errors, aligning with the degraded performance observed.

Users experienced degraded performance as Claude Opus 5 faced elevated errors and unresponsive prompts across Claude Cowork, Claude API (api.anthropic.com), claude.ai, and Claude Code. StatusGator detected the issue at Aug 5, 2026 12:16 PM UTC, roughly 1 hour 35 minutes before the official acknowledgment at Aug 5, 2026 1:52 PM UTC. The early signal underscored a broader impact on Claude's chat capabilities before Claude documented the incident.

Official status updates noted elevated errors and ongoing work to fix the root cause. The outage progressed from degraded performance at 13:52:00Z to continued degradation for several components, with partial recovery starting around Aug 5, 2026 2:18 PM UTC and full recovery by Aug 5, 2026 2:38 PM UTC as all Claude components returned to normal. The incident appeared widespread across Claude services, with users worldwide reporting delays and non-responses.

Users globally experienced degraded performance across Claude Mythos 5, Claude Fable 5, and Claude Opus 5, with prompts slowing, timeouts, and intermittent non-responses. The Claude ecosystem components—Claude Cowork, Claude API (api.anthropic.com), claude.ai, and Claude Code—were affected as the incident unfolded. Impact began around 7:00 AM UTC, and the status page reported degraded performance for the affected models at 7:08 AM UTC as the team identified the cause and worked on a fix. By 9:14 AM UTC, degradation persisted across multiple models while the team continued to address the issue.

An outage began at 8:53 PM UTC with elevated errors across Claude models, affecting multiple components including Claude Cowork, Claude API (api.anthropic.com), claude.ai, and Claude Code. Users experienced issues loading pages, authenticating via OAuth, and making requests to models, resulting in degraded access to Claude services.