← Back to Groq's check readings
We started checking Groq on August 22, 2026, which is 50 days ago. Everything below covers that period, not the full 90 days a longer-tracked service would show.
Email me when Groq reports an incident on its status page, and again when it reports it resolved.
01
The last 24 hours
What happened most recently, on one time axis: our own probe readings by surface, how many people reported a problem, and what Groq's own status page declared. Below it, any public chatter we saw about Groq and the extra probe it triggered.
The last 24 hours on one axis: what our probes read for each surface we check, how many people reported a problem, and what Groq's own status page declared. Three records side by side; none of them changes another.
Our probes, web
Our probes, api (primary)
Unverified user reportsNo reports about Groq in the last 24 hours, which is not evidence that nothing is wrong.
Groq's status pageNo incident declared in the last 24 hours.
24h ago18h ago12h ago6h agonow
Each check cell is the worst reading across every region in that quarter hour: one region that did not answer colours the cell, never an average. An amber cell is a reading that answered, but at three times or more its 7-day median for three checks running: slow is not down, and we say which. A filtered reading stays neutral, because a refused check is not a failed check. Report bars count reports as submitted; we do not check who sent them, and they never change our reading. The status-page lane is Groq's own declaration, shown as they made it.
Third-party chatter, labeled
Public posts from the last hour that mention Groq next to an outage word (“down”, “outage”, “not working”). This is what other people said in public, collected by keyword: not something we measured, and not the user reports above. It never affects the verdict.
No matching public posts in the last hour from the sources we watch.
02
Checks passing by region
Checks passing is the checks that passed, divided by every check we attempted, per region. Not interpolated, not smoothed, and never computed for a region with zero recorded samples.
Their declared record
Groq's own incident history, taken from their public status feed. This is their account of what happened, not our check reading.
Data Center Failure Impacting CapacityMinor incident
Jul 1, 2026, 11:01 PM UTC to Jul 2, 2026, 12:35 AM UTC (1h 34m)resolved
their latest update (Jul 2, 2026, 12:35 AM UTC): The **capacity issue has been resolved**. All services are operating at normal capacity. The incident was caused by a **power loss issue that led to a cooling system failure at one of our US Central data centers**. We apologize for any disruption and **appreciate your patience**.
openai/gpt-oss-120b Performance IssueMinor incident
Mar 19, 2026, 12:51 PM UTC to Mar 19, 2026, 1:50 PM UTC (59 min)resolved
their latest update (Mar 19, 2026, 1:50 PM UTC): The issues affecting openai/gpt-oss-120b have been **resolved**. The model is operating normally. Actions were taken to cancel billing plans and restrict verification status of organizations engaged in coordinated abuse. We apologize for the disruption and thank you for your pati…
meta-llama/llama-4-scout-17b-16e-instruct Degraded PerformanceMinor incident
Feb 7, 2026, 10:02 AM UTC to Feb 7, 2026, 11:34 AM UTC (1h 33m)resolved
their latest update (Feb 7, 2026, 11:34 AM UTC): This incident has been resolved. Thank you for your patience.
meta-llama/llama-4-scout-17b-16e-instruct Degraded PerformanceMinor incident
Feb 5, 2026, 6:09 PM UTC to Feb 5, 2026, 6:32 PM UTC (23 min)resolved
their latest update (Feb 5, 2026, 6:32 PM UTC): The issues affecting meta-llama/llama-4-scout-17b-16e-instruct have been **resolved**. The model is once again operating normally. We apologize for the disruption and thank you for your patience.
meta-llama/llama-4-scout-17b-16e-instruct Degraded PerformanceMinor incident
Feb 5, 2026, 1:23 PM UTC to Feb 5, 2026, 3:07 PM UTC (1h 44m)resolved
their latest update (Feb 5, 2026, 3:07 PM UTC): The issues affecting meta-llama/llama-4-scout-17b-16e-instruct have been fully **resolved**. The model is operating normally. We apologize for the disruption and thank you for your patience.
meta-llama/llama-4-scout-17b-16e-instruct Degraded PerformanceMinor incident
Jan 26, 2026, 11:17 PM UTC to Jan 27, 2026, 1:11 AM UTC (1h 54m)resolved
their latest update (Jan 27, 2026, 1:11 AM UTC): This incident has been resolved. Thank you for your patience.
Data Center Failure Impacting Model Latency - SYDMinor incident
Jan 24, 2026, 4:53 AM UTC to Jan 24, 2026, 7:23 AM UTC (2h 31m)resolved
their latest update (Jan 24, 2026, 7:23 AM UTC): This incident is now resolved.
Model Performance or Availability Issue: llama-3.3-70b-versatileMinor incident
Dec 24, 2025, 11:01 PM UTC to Dec 25, 2025, 12:16 AM UTC (1h 15m)resolved
their latest update (Dec 25, 2025, 12:16 AM UTC): The issues affecting **llama-3.3-70b-versatile** have been **resolved**. The model is operating normally. We apologize for the disruption and thank you for your patience.
Data Center Failure Impacting Capacity - DMM1Minor incident
Dec 17, 2025, 7:11 AM UTC to Dec 14, 2025, 1:30 PM UTC (0 min)resolved
their latest update (Dec 14, 2025, 1:30 PM UTC): **The data center capacity issue has been resolved. All services are operating at normal capacity. The incident was caused by a failed ARP entry on the network gateway device caused the Salam network link to go down in the DMM1 data center, which has been addressed. We apologize …
meta-llama/llama-4-scout-17b-16e-instruct Degraded PerformanceMinor incident
Dec 6, 2025, 6:58 PM UTC to Dec 6, 2025, 8:23 PM UTC (1h 25m)resolved
their latest update (Dec 6, 2025, 8:23 PM UTC): The issues affecting meta-llama/llama-4-scout-17b-16e-instruct , meta-llama/llama-4-maverick-17b-128e-instruct, & groq/compound have been **resolved**. The models are operating normally. We apologize for the disruption and thank you for your patience.
Showing the 10 most recent of 25 incidents Groq has declared in the history we hold.
These entries are Groq's own declarations, taken from their public status feed, shown separately from our own check readings. Their page publishes what they choose to publish: an incident missing here means they did not declare one, not that nothing happened, and an incident here is their account, not our measurement. Neither record ever changes the other.
Groq's own status feed works, so its declared incidents are the record and our probe events are not listed or counted. Our check results by region are above, as a reading beside the statement.
Full detail on what a probe does and does not prove is in the methodology page.