Every status you see on is-down.ai comes from a simple, transparent process designed to be both fast and accurate. Here is the short version. For the full detail, see our methodology.
1. We watch each service continuously
Around the clock, our system checks each AI tool about once a minute, so the status you see is current, not a guess.
2. We check two things at once
- The official status page. Most AI providers publish a status page where they post incidents and maintenance. It is authoritative, but it can lag in the first minutes of an outage, and not every provider offers one.
- An independent reachability check. Separately, we request the service’s public web address and measure how quickly it responds. This catches real problems even when the official page still says everything is fine.
When the two signals disagree, we show the more cautious result and tell you why.
3. We turn that into a plain-language verdict
Instead of jargon, each page tells you clearly whether the tool is up, having problems, or down, with a colour and an icon so it is readable at a glance, plus the response time from our most recent check.
4. We add what only a crowd can tell you
You can report an issue, and each page shows how many people are reporting problems right now. Once enough reports arrive in a short window, we flag that users are affected too. A sudden spike means you are early in an outage; a falling count means recovery is underway. Reports add context, but they never override our own checks.
5. We keep the history
Every check feeds an uptime record, so you can see how reliable each service has been over time, plus a log of recent incidents. We keep detailed records for 60 days and daily uptime summaries indefinitely.