Anthropic outage history

Anthropic Claude users faced a brief degraded experience as several Claude services reported elevated errors. The affected components included Claude Console (platform.claude.com), claude.ai, Claude API (api.anthropic.com), and Claude Code. The incidents began at 6:18 AM UTC and were resolved by 6:33 AM UTC, with the status page indicating a degraded state during the event before services returned to normal.

The recovery was quick, and all listed components returned to standard operation within about 15 minutes. This was a contained incident with no ongoing disruptions after the window closed.

Users attempting to use Claude Sonnet 5 across Claude Cowork, claude.ai, Claude API (api.anthropic.com), and Claude Code experienced degraded performance starting at 3:18 PM UTC. Anthropic noted they were investigating the issue, and status updates described ongoing efforts to restore normal service across theClaude Sonnet 5 suite.

By 3:33 PM UTC, the affected components had returned to normal operation, and Claude Sonnet 5 performance had stabilized. The brief incident primarily affected responsiveness rather than a full outage.

An incident impacted Anthropic Claude services, with error rates rising across multiple models. Users attempting to use Claude Console, Claude Cowork, claude.ai, Claude API, and Claude Code likely faced failed requests or degraded responses during the event. The issue began at 12:53 PM UTC and affected interactions across several major Claude components while investigators worked to diagnose the root cause.

Status page updates confirmed an investigation and restoration. By 1:33 PM UTC all components were back to normal operation. A few user reports surfaced around the incident window, reflecting the disruption experienced by some users.

Users experienced degraded performance on Anthropic Claude Sonnet 5 beginning at 6:18 AM UTC. The incident affected Claude Cowork, claude.ai, Claude API (api.anthropic.com), and Claude Code. Status updates indicated the team was investigating the issue, and services were restored after about 49 minutes, with the incident resolved at 7:08 AM UTC. The episode was brief and did not report a full outage, but it disrupted usage of Claude Sonnet 5 during the window.

Users experienced degraded performance across Anthropic Claude Opus 4.8, affecting Claude Cowork, claude.ai, Claude API (api.anthropic.com), and Claude Code. The incident began at 1:48 PM UTC with an investigation, and users observed slower responses and intermittent failures during the window, lasting about 40 minutes.

Status updates indicated the team was working to restore services, with the affected components returning to normal by 2:28 PM UTC. The impact appeared limited to a subset of users during the outage.

An outage began at 6:03 AM UTC with elevated errors across a broad set of Claude models, including Claude Cowork, claude.ai, the Claude API (api.anthropic.com), and Claude Code. The Claude status page said the team was investigating the issue. By 6:28 AM UTC most models had recovered except Opus 5, which remained affected and was still being recovered. Over the next several hours, recovery moved in fits and starts. At 7:32 AM UTC Opus 5 errors were back to baseline, but Sonnet 5’s errors reemerged; by 8:05 AM UTC Sonnet 5 was stable again, yet Fable 5 showed an elevated error rate. At 8:33 AM UTC the issue spread again across all models, before the service began to stabilize. The Claude components were finally returning to normal by the time the status page noted a full restoration, and the incident was resolved with all components up by 10:08 AM UTC.

An outage affected all Claude models (Claude Cowork, claude.ai, Claude API, and Claude Code), causing elevated errors and degraded responses for users worldwide. The incident began at Jul 29, 2026 7:53 PM UTC with widespread trouble across the Claude suite, and users experienced API errors, timeouts, and overload messages (e.g., 529 Overloaded) during the peak period. The situation evolved through the evening as the provider investigated and worked toward resolution. Recovery began to take hold later, with most models showing improvement before full restoration was announced.

The incident began with elevated errors on Claude Opus 5, affecting Claude Cowork, claude.ai, Claude API (api.anthropic.com), and Claude Code. A follow-up update indicated Haiku 4.5 was also returning errors. The issue was resolved about an hour later, with all listed components returning to normal. The timeline shows investigation underway, then confirmation of broader impact before restoration.

Users experienced elevated errors and degraded performance across the Claude Opus 5 suite, including Claude Cowork, claude.ai, Claude API (api.anthropic.com), and Claude Code. The incident began at Jul 27, 2026 11:28 AM UTC with an active investigation after a user reported issues around 11:15 AM UTC. The degraded period lasted a little over an hour.

By Jul 27, 2026 12:33 PM UTC the Claude Opus 5 services had been restored to normal, with status updates indicating the issue was resolved and services returned to expected operation. The disruption affected multiple components within the Claude Opus 5 suite, impacting workflows that rely on these capabilities.

Users attempting to use Claude Opus 5 encountered elevated errors across multiple Claude services, including Claude Cowork, claude.ai, Claude API (api.anthropic.com), and Claude Code. The official Claude status page indicated an ongoing investigation and listed the affected components as experiencing issues. The incident began when the status changed to warn at 8:18 AM UTC, signaling degraded functionality for affected users. The disruption lasted about 50 minutes and was resolved when the status returned to up at 9:08 AM UTC, with services returning to normal operation.