Provider incident title: “Degraded scale-out due to failed setups pulling from huggingface”
Started
Friday, July 31, 2026
09:25 PM UTC
Duration
2h 14m
total
Resolved
11:39 PM UTC
Jul 31, 2026
Models that pull from huggingface during setup have not been succeeding, which is resulting in scale-out delays and queue backups.
We have confirmed that turning off our caching layer for huggingface downloads allows models to successfully set up, so we are making that change for all models while we continue to troubleshoot.
All affected hardware types have recovered and we are continuing to monitor and troubleshoot. Thanks for your patience!
We're looking stable now, and with caching back on. Thanks again for your patience!
Get alerted next time
Install Tickerr MCP — your agent auto-reports & routes around outages
When Replicate goes down again, your agent reports anonymously and instantly gets a fallback recommendation from the swarm.
Replicate is back online
View current status and uptime history
Replicate 30-day uptime: 100% based on Tickerr's independent monitoring checks.
This incident was sourced from Replicate's official Atlassian Statuspage. Tickerr ingests updates every 10 minutes to surface incidents as they are posted by the provider.
This incident lasted 2h 14m.
Tickerr monitors 90+ AI tools independently. View live Replicate status or all AI tool status.
Get alerted next time Replicate goes down
Tickerr monitors 90+ AI tools. We'll email you when an incident starts or resolves.
Weekly AI pricing & uptime digest
Price drops, new model releases, and incident summaries - every Monday. Free.
Also on Tickerr