LLM API status

OpenAI, Claude & Gemini API status and latency

Is a model API down or slow right now? We send the same short prompt to each model every 5 minutes and time the first token and the whole reply.

Live checks are starting soon

There are no results to show yet. In the meantime, test OpenAI, Anthropic and Gemini with your own API key.

Test LLM API latency

Frequently asked questions

Is the OpenAI API down right now?
The OpenAI card above shows the result of our latest check, made in the last few minutes, and how many checks failed in the last 24 hours. If it's answering here but not for you, check your API key, your rate limits and your account's credits.
What is time to first token?
The time from sending a request to receiving the first visible piece of the reply, with streaming on. It's how long people wait before text starts to appear in a chat app, so it usually matters more than the total time.
How are these numbers measured?
Every 5 minutes our server sends the same short prompt to each model, with streaming on, and times the first token and the end of the reply. Medians and the 95th percentile are over the checks that replied in the last 24 hours.
Why are my own latencies different?
Your prompts are longer, your region and network differ, and providers route accounts differently. Test your own key and model with the free LLM API latency tester, or monitor them around the clock.

Know when your model fails

These checks use our keys. Monitor the model your product uses, on your own key, and get alerted by email, Slack, SMS or a phone call when it stops replying or slows down.

Monitor your LLM API free
Test your own key and model now