Anthropic incident: Degraded performance for Claude Opus 5 and Claude Haiku 4.5

Affected components

• Claude API (api.anthropic.com)
• Claude Code
• Claude Cowork
• claude.ai

1. resolvedAug 19, 2026, 11:02 UTC

   This incident has been resolved.

2. monitoringAug 19, 2026, 10:33 UTC

   A fix has been implemented and we are monitoring the results.

3. identifiedAug 19, 2026, 09:42 UTC

   We have identified the cause of elevated errors on requests to Claude Opus 5 and Claude Haiku 4.5 and are working on a fix. We will provide an update as soon as possible.

What APIStatus.watch saw

Our 5-minute checks didn't record a change in Anthropic's overall status around this incident. Smaller or regional incidents often leave a provider's overall status green.

Other services with incidents at the same time

Overlapping incidents aren't necessarily related.

• Elasticsearch: Elasticsearch 9.5.1: false-positive matches in certain boolean queriesMajor · Aug 14, 2026, 09:22 UTC
• Supabase: 401 errors due to JWT rejectionsMinor · Aug 14, 2026, 02:23 UTC
• Telnyx: Programmable Wireless data loss in AustraliaMinor · Aug 17, 2026, 19:27 UTC
• Resend: Contact webhook events are delayed in deliveryMinor · Aug 18, 2026, 19:53 UTC
• Twilio: SMS Delivery Receipt Delays from Twilio to Telecall BrazilMinor · Aug 19, 2026, 06:08 UTC
• Twilio: SMS Delivery Delays from Twilio to Telenet BidCo NV and Base Networks in BelgiumMinor · Aug 19, 2026, 07:50 UTC
• Twilio: SMS Delivery Failures from a Subset of Twilio Long Codes to Bell Aliant CanadaMinor · Aug 19, 2026, 09:40 UTC

How often does Anthropic have incidents?

Anthropic reported 85 incidents in the last 90 days, 32 of them major or critical. A typical incident lasted 1h 7m. See Anthropic's uptime and incident history