Status for models, routes, and providers.
Track availability, latency, and recent route incidents. This page is meant to become a trusted reference when teams choose what feels safe enough for production.
Monitored Routes
9
provider + alias pairs
Healthy Routes
6
67% currently green
Degraded Routes
3
watching for recovery
Median Latency
581ms
cross-route p50
GPT-4o via OpenAI
healthy98.72%
99.41% availability · 416 / 1488ms
Availability
99.41%
p50 latency
416ms
p95 latency
1488ms
Throughput
44 t/s
Time window
24h
GPT-4o fallback via DeepSeek
degraded92.11%
97.90% availability · 851 / 2390ms
Availability
97.90%
p50 latency
851ms
p95 latency
2390ms
Throughput
37 t/s
Time window
24h
GPT-4o Mini via OpenAI
healthy97.94%
99.12% availability · 268 / 811ms
Availability
99.12%
p50 latency
268ms
p95 latency
811ms
Throughput
68 t/s
Time window
24h
Claude Sonnet 4 via Anthropic
healthy96.84%
98.87% availability · 734 / 1846ms
Availability
98.87%
p50 latency
734ms
p95 latency
1846ms
Throughput
31 t/s
Time window
24h
Claude Sonnet 4 fallback via OpenAI
degraded91.43%
96.53% availability · 692 / 2105ms
Availability
96.53%
p50 latency
692ms
p95 latency
2105ms
Throughput
28 t/s
Time window
24h
Claude 3.5 Haiku via Anthropic
healthy97.08%
99.02% availability · 318 / 906ms
Availability
99.02%
p50 latency
318ms
p95 latency
906ms
Throughput
59 t/s
Time window
24h
DeepSeek Chat via DeepSeek
healthy95.63%
98.31% availability · 521 / 1673ms
Availability
98.31%
p50 latency
521ms
p95 latency
1673ms
Throughput
52 t/s
Time window
24h
DeepSeek Chat fallback via OpenAI
healthy94.88%
97.65% availability · 402 / 1322ms
Availability
97.65%
p50 latency
402ms
p95 latency
1322ms
Throughput
61 t/s
Time window
24h
DeepSeek Reasoner via DeepSeek
degraded90.71%
95.94% availability · 1024 / 3088ms
Availability
95.94%
p50 latency
1024ms
p95 latency
3088ms
Throughput
21 t/s
Time window
24h