Pulse
Grok uptime and downtime by month
Grok uptime and downtime per component, month by month — counted only inside recorded incidents, never as an aggregate service figure.
How much downtime has Grok had this month?
October 2026 is still open: 2 recorded incidents so far, counted per component. The figure moves until the month closes.
- October 2026Month to date — still moving until the month closes2 incidents · 10 components
- September 2026Complete month10 incidents · 10 components
- August 2026Complete month0 incidents · 10 components
- July 2026Complete month1 incidents · 10 components
- June 2026Complete month0 incidents · 10 components
- May 2026Complete month0 incidents · 10 components
- April 2026Complete month0 incidents · 10 components
- March 2026Complete month7 incidents · 10 components
- February 2026Complete month6 incidents · 10 components
- January 2026Complete month23 incidents · 10 components
- December 2025Complete month1 incidents · 10 components
- November 2025Complete month1 incidents · 10 components
- October 2025Complete month2 incidents · 10 components
- September 2025Complete month0 incidents · 10 components
- August 2025Complete month2 incidents · 10 components
- July 2025Complete month0 incidents · 10 components
- June 2025Complete month0 incidents · 10 components
- May 2025Complete month2 incidents · 10 components
- April 2025Complete month5 incidents · 10 components
- March 2025Complete month4 incidents · 10 components
- February 2025Complete month1 incidents · 10 components
Coverage line: the first incident in SpaceXAI's own feed is dated January 15, 2025, so the first month we can measure end to end is February 1, 2025. January 2025 has no page — a part-month figure would look exact and be false. Every figure here is derived from recorded incidents only, never from the live component states.
How this is counted
- Source. SpaceXAI's public incident reporting for Grok. We do not probe, ping or measure the services ourselves, so nothing here is an availability measurement.
- What we count. Minutes of the month that fell inside a recorded incident window for that component — from the reported start to the reported resolution, clipped to the month.
- What we do not count. Severity is ignored: an incident logged as degraded performance counts the same as a full outage. Unreported disruption, regional differences and per-user impact are invisible to us. An incident window is not the same thing as downtime.
- Overlaps. Figures are per component only, never aggregated into one service-level number. Within a component, overlapping incidents count the shared minutes once; the same minute can appear under several components when an incident affected several.