Disruption with Claude Opus 4.7
This summary is created by Generative AI and may differ from the actual content.
Overview
On June 8, 2026, from 08:40 UTC to 09:30 UTC, the Claude Opus 4.7 model used in Copilot Chat, VS Code, and other Copilot products experienced degraded availability, with error rates peaking at 8.4 % and averaging 1.9 %. The degradation was traced to an upstream model‑provider issue that caused temporary unavailability and rate‑limiting on secondary fail‑over systems. Users who selected the Auto option or alternative models were not affected.
Impact
The incident reduced the availability of the Claude Opus 4.7 model, resulting in elevated error rates (peak 8.4 %, average 1.9 %) for users of Copilot Chat, VS Code, and related Copilot surfaces during the 50‑minute window. No information on user‑impact magnitude or revenue impact was provided.
Trigger
An upstream model‑provider outage introduced temporary unavailability and imposed rate limits on the secondary fail‑over paths that serve Claude Opus 4.7.
Detection
Degraded performance was identified through internal reports and monitoring that flagged increased error rates for the Claude Opus 4.7 model, prompting an investigation that began at 09:05 UTC.
Resolution
The team coordinated with the upstream provider to restore service, observed recovery as rate limits were lifted, and advised customers to switch to alternative models or the Auto setting. Ongoing work includes enhancing provider fail‑over mechanisms and improving monitoring to prevent recurrence.
Root Cause
The root cause was the upstream model‑provider’s outage, which caused temporary loss of service and rate‑limiting on the secondary fail‑over systems that supply Claude Opus 4.7.