Free website status API
Live status, uptime and outage history for all 179 monitored services, as JSON. Free, with 1,000 requests a day - You just need a key, and a key needs a domain you have verified.
Getting a key
Proving you own the domain
Three steps, the same way AdSense verifies a site. About two minutes, and it never needs repeating.
- Sign in with Google and claim a domain you control.
- Prove it. Add a DNS TXT record, or put a small file at /isitdown-verify.txt on the domain.
- Press verify. Checked on the spot, and the key is issued immediately.
Why the extra step? This API used to be completely open, which is pleasant right up until one misconfigured loop serves more requests than every human visitor put together - And there is no way to contact whoever wrote it. Tying a key to a verified domain means an abusive key has a real owner we can email rather than a subnet we have to block. That is what lets the free tier stay genuinely free.
Limits
Per-key quotas
The API never contacts the sites it describes. It returns measurements we have already taken on our own schedule, so no number of API calls sends a single extra request to anyone's server.
Quotas are per account - shared by every key you hold, including ones revoked today - and reset at 00:00 UTC. Every response carries X-RateLimit-Used, X-RateLimit-Limit and X-RateLimit-Remaining, so a well-behaved client never has to guess.
Endpoints
Send the key as a bearer token: Authorization: Bearer idk_…
GET/api/v1/status/<slug>
Everything about one service. Add /detail to the path for the 24-hour report series, the problem breakdown and the last ten incidents.
GET/api/v1/services
Every tracked service with its most recent check. Filter by category in the path: /api/v1/services/social.
GET/api/v1/down
Only the services currently failing their check, with how long each has been down. The endpoint to poll for a dashboard.
GET/api/v1/trending
Services attracting the most user reports in the last six hours.
GET/api/v1/summary
One small object: How many services are tracked, how many are down, and the 24-hour check and report totals. This one needs no key - It is the endpoint to try before you sign up, within 60 requests an hour per address.
Response shape
Fields every response carries
{
"slug": "facebook",
"name": "Facebook",
"domain": "facebook.com",
"status": "up",
"status_text": "No problems detected",
"checked_at": "2026-09-19T08:00:11Z",
"response_ms": 421,
"http_status": 200,
"reports_24h": 3,
"uptime_30d": 99.986,
"downtime_30d_seconds": 372,
"current_outage": null
}status is one of up, slow, down, reports, checking or unknown. Only down means an outage our checks confirmed; reports means the service answered us but users are complaining.
Terms, briefly
Use it in anything, commercial or not. Cache responses for at least 60 seconds - They do not change faster than that. If you build something public with it, a link back is appreciated and never required.
What you should not do: Poll harder than the rate limit, present this data as a guarantee of anything, or re-publish it as if it were the provider's own status. It is one network's view, and the methodology page is honest about the limits of that.
There is no uptime commitment on the API itself, which would be a strange thing for a site about outages to pretend to offer.