Anthropic Reports Degraded Performance Across All Claude Models

Starting 16:20 UTC on Aug 18, Anthropic's status page reported degraded performance across multiple Claude models, marking the incident as under investigation and promising updates via status.claude.com. As of posting, no root cause or fix progress has been disclosed.

Confirmed

• Affected models include Mythos 5, Fable 5, Opus 5, Sonnet 5 and Haiku 4.5 — the entire model lineup
• Affected services include the Claude app (web), API, Claude Code and Cowork; Anthropic separately confirmed elevated error rates for Opus 5 and Haiku 4.5 requests
• Monitoring site Status Is Down and user reports indicate hundreds of users unable to access the service, spanning web and desktop apps

Unconfirmed

• No official explanation of root cause or remediation. @thedealdirector relayed user reports of degraded response quality; pending status page updates

Why it matters

• A simultaneous degradation across the entire model lineup is rare for Anthropic and directly impacts developer API usage and production toolchains such as Claude Code

2026-08-19 ~ 2026-08-19 · 6 related posts

• [source] Anthropic investigating degraded performance across multiple Claude models — ClaudeAI-mod-bot · 2026-08-19
• [source] Claude and Anthropic AI experiencing major outage affecting hundreds of users — johnmccrea · 2026-08-19
• Anthropic Experiencing Widespread Outage Across All Models — ns123abc · 2026-08-19
• [source] Anthropic reports degraded performance for Claude Opus 5 and Haiku 4.5 — ClaudeAI-mod-bot · 2026-08-19

2 near-duplicate retellings: ns123abc · thedealdirector