Cursor incident: Investigating service degradation

Affected components

• Automations
• CLI
• Cloud Agents
• IDE

1. resolvedSep 2, 2026, 03:07 UTC

   This incident has been resolved.

2. monitoringSep 2, 2026, 01:00 UTC

   A fix has been implemented and we are monitoring for service restoration.

3. investigatingSep 2, 2026, 00:48 UTC

   We're currently investigating degraded performance affecting requests made with Composer 2.5 and Grok 4.6 Fast. While these models are in use, you may encounter failed or retried requests across Automations, Cloud Agents, the CLI, and the IDE. Other models are unaffected.

What APIStatus.watch saw

Our 5-minute checks didn't record a change in Cursor'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: Replication bug causing slow recoveriesMajor · Sep 1, 2026, 08:16 UTC
• OVHcloud: [EU-SOUTH-MIL][Containers & Orchestration] - Managed Private Registry incident notificationMinor · Aug 26, 2026, 15:00 UTC
• Snowflake: INC20000150Minor · Aug 28, 2026, 00:07 UTC
• Snyk: Partial Broker Service DegradationMinor · Aug 30, 2026, 14:32 UTC
• Anthropic: Delays in credit purchasesMinor · Sep 1, 2026, 23:26 UTC
• Datadog: Delayed DBM Monitor EvaluationMinor · Sep 2, 2026, 01:07 UTC

How often does Cursor have incidents?

Cursor reported 50 incidents in the last 90 days, 15 of them major or critical. A typical incident lasted 1h 8m. See Cursor's uptime and incident history