Model API status / OpenRouter
API endpoint for a multi-model aggregation and routing service.
Probed from Cloudflare edge DFW · updated 2 min ago (10/17 19:45)
Measured in mainland China
| API | All | China Unicom | Mobile |
|---|---|---|---|
| OpenRouter11 samples | 306 ms成功 100% | 527 ms成功 100% | Insufficient samples |
Values are the median round-trip time over a reused connection. “Success” is the share of requests that received a response within 6 seconds.
FAQ
As of 10/17 19:45, openrouter.ai is reachable from Cloudflare edge nodes with a response time of 71 ms. Over the past 24 hours, reachability was 100% and median latency was 66 ms.
过去 7 天中国大陆访客共测得 11 个样本，请求成功率 100%，连接复用后往返中位数 306 ms，首次连接中位数 761 ms。
Every 5 minutes, we send a request without an API key from Cloudflare edge nodes to https://openrouter.ai/api/v1/auth/key. Any status code below 5xx, such as 401, counts as reachable. We record the time from sending the request to receiving the response headers. See the methodology page for details.
Related articles
The same model can swing between good and bad because inference providers quantize differently. We count fp8, fp4 and int4 across 137 OpenRouter endpoints and show how to pin precision.
Compare 1-day uptime and quantization for the same model across OpenRouter providers using a 2026-10-17 cdnping snapshot, learn how to read nulls, and see how to set provider preferences.
How to track LLM API price changes, new models and deprecations: three sources to watch, cdnping's /changes RSS and /models, and team alerts.