Elevated GPU provisioning latency
MinorResolved
Resolved. Upstream provisioning latency returned to normal.
Identified upstream congestion in the GPU provisioning API.
Investigating reports of slow GPU provisioning.
Live platform health
System status
Production routes, auth, database, and GPU provisioning are responding normally.
All Systems Operational
No active incidents- Services
- 7/7
- GPUs
- 6/6
- 90 days
- 96.67%
Services
Production routes and dependencies
API Gateway
healthy
GPU Provisioning
HTTP 200
Session Management
auth gate responding
Authentication
auth gate responding
Database
ok
Stripe API
Supabase Auth
GPU availability
Current launchable capacity
GPUVRAMStatus$/hr
RTX 4090
24GB1 units$0.79RTX 5090
32GB1 units$1.14RTX 6000 Ada
48GB2 units$0.89H100 SXM 80GB
80GB8 units$3.78H200 141GB
141GB8 units$5.05B200 180GB
180GB4 units$6.7790-day uptime
87 clear days of 90
90 days agoToday
Subscribe to updates
Incident updates
Get notified when services go down or incidents are resolved.
Incident history
Public incident ledger
Session API intermittent 500s
MajorResolved
Resolved. Edge runtime migration fixed the underlying issue.
Identified: edge→node migration required for session route.
Investigating elevated error rate on /api/sessions.
Cold start times exceeding 60s for large images
MinorResolved
Resolved. Image layer caching reduced cold starts to <15s.
Investigating: 7GB Docker image causing long cold starts.