Anthropic Outage History

From Sep 27, 2025 11:45 PM UTC to Sep 28, 2025 2:17 AM UTC, Anthropic Claude users experienced failures loading conversations, with errors such as 'conversation not found' reported across sessions and devices. StatusGator detected the issue at Sep 27, 2025 11:45 PM UTC ahead of any official acknowledgement, highlighting an Early Warning Signal. The outage disrupted core chat functionality for several hours and prevented users from starting or continuing conversations.

Anthropic's Claude API (api.anthropic.com) experienced elevated errors starting at Sep 27, 2025 5:29 PM UTC, resulting in degraded performance for users calling Opus 4 and Sonnet 3.5 endpoints. The incident affected Claude API users relying on these endpoints and prompted ongoing investigation.

The status page issued two updates during the incident, and the service was resolved by 10:23 PM UTC. The outage lasted approximately 4 hours and 54 minutes.

Claude services experienced a brief degradation starting at Sep 25, 2025 2:08 PM UTC, with elevated errors on Claude Sonnet 3.5 and 3.7. The incident affected four components: claude.ai, Claude API (api.anthropic.com), Claude Code, and platform.claude.com, likely impacting API usage, code features, and console access. The outage lasted about 3 hours and 25 minutes and was resolved by Sep 25, 2025 5:33 PM UTC. The provider issued two status updates during the event, reflecting an initial investigation and a restoration to normal operation.

Anthropic Claude Sonnet 3.7 experienced a partial outage affecting two components: the Claude API (api.anthropic.com) and platform.claude.com. Users attempting API requests or using the web console may have encountered errors or degraded performance during the incident. The outage began at Sep 25, 2025 12:52 AM UTC and was listed as degraded at Sep 25, 2025 12:53 AM UTC, lasting about 2 hours and 16 minutes until it was resolved by Sep 25, 2025 3:08 AM UTC.

The provider issued two status updates during the incident: the first noting investigation and elevated errors, and the second confirming restoration to normal. The outage affected Claude API and platform.claude.com throughout the event.

Anthropic reported elevated error rates across Claude API (api.anthropic.com), claude.ai, Claude Code, and the Claude platform (platform.claude.com), resulting in degraded performance for users attempting to access Claude AI features or make API calls. Users may have experienced failed requests or slower responses while the incident was active, with four components affected: claude.ai, Claude API, Claude Code, and the Claude platform.

Anthropic Claude experienced degraded performance beginning at Sep 24, 2025 8:03 AM UTC as Claude Opus 4 and Opus 4.1 on Claude Code reported elevated errors. This affected Claude AI interfaces and API, including claude.ai, the Claude API (api.anthropic.com), and the platform portal (platform.claude.com). Users encountered API errors and overload messages, indicating a broad impact on developers and users relying on Claude services.

The Claude status page published three updates during the incident: initial investigation at Sep 24, 2025 8:03 AM UTC, a note that a fix had been implemented and that monitoring results were being observed at Sep 24, 2025 8:53 AM UTC, and a resolution at Sep 24, 2025 9:13 AM UTC stating that services had returned to normal. The outage lasted approximately 1 hour and 10 minutes. A noticeable surge in user reports around the incident reflected widespread impact across the Claude API and related interfaces, with reports from multiple regions.

Anthropic's Claude API experienced degraded performance with elevated error rates when fetching images and PDFs by URL on the Claude API (api.anthropic.com). This affected applications relying on image or PDF retrieval through the API. The incident began at Sep 23, 2025 6:32 PM UTC, with the status page noting investigation and a transition to a degraded state at Sep 23, 2025 6:33 PM UTC, and was resolved by Sep 23, 2025 8:18 PM UTC, lasting approximately 1 hour and 46 minutes. The affected component is the Claude API (api.anthropic.com).

Claude App and API connectivity issues affected four components: platform.claude.com, claude.ai, Claude API (api.anthropic.com), and Claude Code. The outage began at Sep 22, 2025 9:03 PM UTC and lasted until Sep 22, 2025 9:13 PM UTC, with degradation to some endpoints noted by Sep 22, 2025 9:08 PM UTC. The incident prompted 3 official updates from Anthropic as the status moved from down to degraded to up. User reports started around Sep 22, 2025 8:45 PM UTC and continued through the incident, indicating a broad but brief impact across users.

The incident began around Sep 22, 2025 7:41 AM UTC with elevated error rates affecting Anthropic Claude API, claude.ai, Claude Code, and platform.claude.com. Users attempting to interact with Claude services experienced errors or unresponsive responses during this period. A status update indicated the issue was under investigation, and the problem persisted across multiple components (platform.claude.com, Claude API at api.anthropic.com, Claude Code, and claude.ai). A user reported Claude interactions appearing unresponsive. The outage lasted approximately 1 hour and 15 minutes, with the provider issuing 2 status updates as the situation progressed, and services returned to normal by Sep 22, 2025 8:58 AM UTC.

Anthropic Claude Haiku 3 and Claude Haiku 3.5 experienced elevated errors, impacting user calls to Claude AI and the Claude API (api.anthropic.com). Haiku 3 showed a very high error rate from Sep 20, 2025 2:00 AM UTC to Sep 20, 2025 5:00 AM UTC, and Haiku 3.5 was affected from Sep 20, 2025 2:30 AM UTC to Sep 20, 2025 3:00 AM UTC. This caused degraded performance for model usage and API requests during those windows. The two affected components were claude.ai and Claude API (api.anthropic.com).