Requests are proxied to the nearest available provider. Traffic fails over automatically if an upstream degrades or times out.
MODELPROVIDERSTATUS
GPT-4.5
OpenAI
operational128ms
Claude Opus 4.8
Anthropic
operational154ms
Claude Sonnet 5
Anthropic
operational96ms
Gemini 2.5 Pro
Google
degraded310ms
Llama 4 Maverick
Meta
operational88ms
Mistral Large 2
Mistral AI
operational121ms
DeepSeek V3
DeepSeek
operational142ms
Grok 3
xAI
operational176ms