Cursor outage on September 21, 2026: Investigating service degradation
Cursor reported an incident affecting Cursor for 22 minutes on 21 September 2026, from 06:39 to 07:00 UTC.
Resolved
Official incident reported by Cursor
- Started
- 06:39
- 21 Sep 2026 UTC
- Resolved
- 07:00
- 21 Sep 2026 UTC
- Duration
- 22m
- 22 minutes
- Impact
- Minor
- as rated by Cursor
- Updates
- 2
- posted by Cursor
Timeline
All times UTC.
Official incident from Cursor
Investigating service degradation
-
Official
Investigating
We are investigating degraded performance for requests that use Grok 4.6. Users may see failed or retried requests in Automations, Cloud Agents, the CLI, and the IDE when Grok 4.6 is used. Other models are unaffected.
-
Official
Resolved
This incident has been resolved.
"Official" rows are quoted from Cursor's public status page, status.cursor.com. Very long posts are shortened.
What our checkpoints saw
HostTracker's checkpoints, which test cursor.com from several locations, recorded no outage of 5 minutes or longer during this incident. A provider can report a problem that affects only some features, regions or accounts, which a check of the main address does not see. The live Cursor status shows what our checks see right now.
Does your product depend on Cursor?
An outage like this one reaches your users through your product, and most teams hear about it from a customer rather than from a dashboard. HostTracker watches any endpoint you name - your own site, an API you call, a vendor you resell - from 300+ locations, confirms a failure from several of them before it says anything, and then tells you by email, SMS, voice call or your chat tool.
You get the same record you are reading here for your own services: every outage timestamped, with the checks that observed it.
Independent measurement by HostTracker. Not affiliated with Cursor. Official updates are quoted from Cursor's public status page and remain Cursor's own words.