Anthropic Outage History

Anthropic Claude encountered elevated errors across multiple models beginning at Sep 18, 2025 7:35 PM UTC. The issue impacted four components: platform.claude.com, claude.ai, Claude API (api.anthropic.com), and Claude Code, causing degraded or partially functioning responses for users. The status page indicated investigation started as the incident emerged.

Anthropic Claude Sonnet 4 experienced elevated errors starting at Sep 16, 2025 6:35 AM UTC, affecting Claude.ai, Claude Code, console.anthropic.com, and api.anthropic.com. Users attempting to use Claude Sonnet 4 encountered errors or degraded responses during this period. The Claude status page listed these four components as degraded around Sep 16, 2025 7:03 AM UTC while the team investigated the issue.

By Sep 16, 2025 8:33 AM UTC, Claude Sonnet 4 services had returned to normal. The incident lasted roughly 1 hour and 58 minutes from start to finish, and the provider issued status updates as the situation progressed.

User reports were limited, with a small volume describing errors or degraded responses across the affected Claude components.

The outage began at Sep 16, 2025 12:57 AM UTC and was resolved by Sep 16, 2025 1:28 AM UTC. It affected the Claude.ai Artifacts loading feature, with users potentially encountering errors when loading Artifacts during the incident window. The Claude.ai status was briefly downgraded to a degraded state, and the incident lasted about 31 minutes. The provider posted two status updates: at Sep 16, 2025 12:58 AM UTC (Investigating) and at Sep 16, 2025 1:28 AM UTC (Restored).

Anthropic's Claude Haiku 3.5 experienced a brief degradation beginning at 11:13 AM UTC, affecting Claude.ai, Claude Code, console.anthropic.com, and api.anthropic.com. Users may have encountered errors when using Claude Haiku 3.5 features. The incident lasted about 1 hour 15 minutes and was resolved by 12:28 PM UTC. The Claude status page reported two updates during the event, including a clarification that the affected model was Claude Haiku 3.5 rather than Claude Opus 4.1.

Anthropic Claude Opus 4.1 experienced a brief degradation affecting four core components: claude.ai, Claude Code, console.anthropic.com, and api.anthropic.com. The issue began around 6:15 AM UTC and the Claude status page indicated elevated errors at Sep 15, 2025 7:28 AM UTC, with users potentially seeing errors or degraded functionality for a few hours. The incident lasted for roughly 9:18 AM UTC and was resolved by that time.

Anthropic Claude experienced a brief degradation in model output quality affecting multiple endpoints: console.anthropic.com, api.anthropic.com, Claude Code, and claude.ai. Users likely encountered lower quality outputs or slower responses during the incident across these components.

Anthropic Claude Opus 4.1 experienced elevated errors affecting multiple components: api.anthropic.com, Claude Code, console.anthropic.com, and claude.ai. Users attempting to access the API, use Claude Code features, or interact with Claude via the UI may have encountered errors or degraded performance. The issue began around Sep 11, 2025 1:10 PM UTC and was acknowledged as a degraded service from Sep 11, 2025 1:28 PM UTC. The service was restored by Sep 11, 2025 2:33 PM UTC, with two status updates issued during the incident. The outage lasted about 1 hour and 5 minutes.

On 2025-09-10, Anthropic Claude experienced a global outage affecting api.anthropic.com, claude.ai, and console.anthropic.com. The outage began at 4:33 PM UTC and lasted about 1 hour and 5 minutes, with users unable to access the Claude API, Claude.ai features, or the Console. The incident involved three components and disrupted typical Claude workflows across services relying on Claude.

Anthropic released five official status updates during the incident, including a mid-event fix and ongoing monitoring. By 5:38 PM UTC, all affected components had returned to normal, marking the end of the outage.

From 12:18 AM UTC to 9:23 AM UTC, Anthropic Claude experienced degraded output quality across multiple components, affecting API responses, Claude Code, console, and claude.ai. Users reported lower-quality model results and, at times, elevated errors during this window. The incident lasted about 33 hours and involved four components listed in the status updates.

During the incident, the Claude status page issued four official updates documenting the degradation, investigation, and remediation efforts. The service flagged degraded model output quality, then later reported elevated errors on Claude Sonnet 4.0 as remediation progressed. The team rolled out fixes and continued monitoring until the issues were resolved.

Impact and scope: The outage affected API access at api.anthropic.com, Claude Code, console.anthropic.com, and claude.ai. Reports from users in the United States, United Kingdom, and India indicated broad, cross-regional impact rather than a single geographic pattern. The incident concluded with the service restored and monitoring ongoing as of 9:23 AM UTC.

Anthropic's Claude platform experienced a brief degraded state impacting messaging to Opus 4.1, affecting claude.ai, console.anthropic.com, api.anthropic.com, and Claude Code. Users attempting to send messages to Opus 4.1 could experience failures during the incident window. The provider indicated an investigation and moved the affected components to a degraded status.