Recent incidents (last 90 days)
- A100 hardware partial outage2026-08-26 · 3 hmajor
- Delayed scaling due to node failure2026-08-10 · 40 minminor
- Degraded scale-out due to failed setups pulling from huggingface2026-07-31 · 2 hminor
- Hitting GPU Capacity for H100s creating large queue times for some models2026-07-22 · 26 hminor
- H100 GPU shortage resulting in high queue times2026-07-13 · 25 hminor
- High contention on H100 hardware2026-07-10 · 8 hminor
- Limited H100 capacity2026-06-30 · 6 hminor
How we know
Datalenk polls the official Replicate status page every fifteen minutes and keeps the history here.
See how Replicate incidents show up on your own traffic.
When it matters to your revenue, you should not have to guess.
Explore Datalenk →