All systems operational
Uptime 99.9% over the last 90 days. Last incident: September 27 — elevated latency for 40 minutes, resolved.
| Component | Status |
|---|---|
Messages API /v1/messages | Operational |
Chat Completions /v1/chat/completions | Operational |
| Dashboard and sign-in | Operational |
GLM 5.3 glm-5.3 | Operational |
DeepSeek V4 Pro deepseek-v4-pro | Operational |
KIMI K3.1 Pro kimi-k3.1-pro | Operational |
Median time to first token. Higher bars are busier hours; the pool is shared, so peak hours are slower for everyone.
Trial plans share one concurrency pool. Getting
429 rate_limit_error during peak hours is expected on the free plan
and is not a problem with your client — retry, or move to a paid plan for
priority. See the docs.