Modal had 36 outages in the last 2 years totaling 36h 49m of downtime — averaging 1.5 incidents per month.
There were 36 Modal outages since December 28, 2025 totaling 36h 49m of downtime. Each is summarised below — incident details, duration, and resolution information.
Task logs have been delayed since 2:46 PM ET
We have identified and fixed a bottleneck; task logs are flowing through the system normally
We have identified and fixed a bottleneck; task logs are flowing through the system normally
We're looking into an issue scheduling containers
We have recovered
We have recovered
We have identified the root cause and are taking action to mitigate.
All issues have been resolved.
We are working on recovery.
All other input planes are now recovered.
Due to a Redis cluster node becoming saturated Function input fetching and output publishing has an elevated failure rate. We're actively working on remediation to bring down error rates to zero.
Error rate is back to zero and service is fully restored.
Our observability data storage provider is having an outage. Dashboard is severely degraded for all users, function execution should not be affected.
The dashboard is recovered.
Our storage backend began bandwidth throttling us, resulting in some errors / degraded Modal Volume performance.
We mitigated and are no longer getting bandwidth throttled.
An expired TLS cert slipped past our early alerting causing the Flash Proxy routing service to go down in us-west only.
We refreshed the cert and the service has recovered.
We are experiencing intermittent issues with accessing Volumes via CLI, API and the Modal dashboard. Access to Volumes from containers is not affected.
The issue has been mitigated and normal access to Volumes via CLI, API and the dashboard is restored.
We experienced a drop in capacity across our fleet.
All capacity is back.
CPU functions, GPU functions, Snapshot restores (beta), and 2 other resources went down.
All services operational as of 11:06 UTC.