Degraded Performance Agent SDK
The Agent SDK has degraded performance. This affects Grok Build and Cursor. xAI is investigating.
State
Investigating
Started
Ongoing for
20 min
Worst impact
Degraded
Timeline
1 updateInvestigating
20 min 100%
Updates
Times in UTCAs posted by xAI
We are investigating degraded performance affecting the Agent SDK.
Affected components
1xAI in the last 90 days
Full historyIncidents in 90 days
59
Days with an incident
37
Median duration
58 min
This incident
65% shorter than the median
Other xAI incidents
All incidentsxAI·Degraded Gemini 4 Carbon
The Gemini 4 Carbon model on Third-party Model Providers does not work correctly. Some requests fail intermittently. xAI reports that the status is investigating. xAI continues to work with an upstream provider on a fix.
xAI·Elevated errors for some requests
xAI reports elevated errors for some requests. The incident affects cursor.com and Grok Bot. xAI resolved the incident after 91 minutes.
xAI·Usage data slow or unavailable for some users.
On 10 November 2026, usage data on the grok.com dashboard and usage APIs was slow or showed no usage for some users. xAI reported the incident as investigating. xAI then reported the incident as resolved. The incident lasted 1 hour and 3 minutes.
xAI·Elevated errors for OpenAI model requests
xAI reported elevated errors for requests to OpenAI models. The issue affected Automations, Cloud Agents, CLI, Cursor IDE (Desktop), and third-party models. xAI resolved the incident after 34 minutes.
xAI·Grok Bot is currently degraded
Grok Bot is degraded. xAI reports that the incident is resolved. The outage lasted 1 hour and 17 minutes.
Follow xAI
Your coding agent knows when xAI is down, and tells you when a failure is theirs, not yours. You can also get xAI incidents in a feed reader or in Slack.
Install the outages.live MCP server, so that when something breaks you can tell if the cause is a provider we depend on and not my code. It reads the official status pages of cloud, AI, payment and developer providers and gives you their incidents as MCP tools.
Add it as a remote MCP server named "outages" at https://outages.live/mcp (Streamable HTTP, no authentication), with your CLI command or your MCP config file.
Then:
1. Make sure it connects and shows the tools get_provider, list_feed and send_feedback.
2. Call get_provider with "xai" and tell me if xAI has a problem right now.
3. From now on, when a call to a third-party service fails or becomes slow, call get_provider for that service before you change my code.Paste this into any coding agent. It adds the server with its own MCP setup.
https://outages.live/feed.xml?providers=xai/feed subscribe https://outages.live/feed.xml?providers=xaiFAQ
Go from tracking outages to fixing production issues with Polylane.
Try Polylane