Undownable checks your websites, APIs, and servers around the clock (HTTP, TCP, ping, DNS, and SSL) and alerts you the moment something breaks. Outages are confirmed across repeated checks before you're paged, so a blip never wakes you; public status pages keep your users in the loop.
Start with a 14-day Pro trial, then free forever for 10 monitors. No card required.
Real numbers, across every team on Undownable, updated hourly.
This is the same cell view you get inside Undownable, watching real endpoints right now. They live on downable.site, our chaos target, and go down on a fixed schedule. Green is up, red is an outage Undownable would have caught.
Real state, pulled live from downable.site, our chaos target: it fails on a fixed schedule so these cells always have something honest to show.
Your dashboard opens with this grid: ninety days of checks, one cell per four hours. This one is the real thing, aggregated and anonymized across every monitor on Undownable. Yours shows only your own.
HTTP(S), keyword, TCP port, ping, DNS records, SSL certificate expiry, and push heartbeats. One dashboard for all of it.
Confirmation thresholds require consecutive failures before a monitor is marked down, so a single blip never pages you.
Down to 1-minute intervals on Pro. Incidents open and resolve automatically with full timelines.
Branded, always-current status pages with uptime history for your plan: up to a year, on your own custom domain, with Pro.
Email, ntfy, Telegram, Slack, Discord, and webhook notifications the moment an incident opens or resolves.
Exactly-once delivery, automatic retries, and dead channels that disable themselves. Flap suppression pages you once, not ten times.
Undownable ships a first-party MCP server (Model Context Protocol), so the AI agents you already build can create monitors, read incidents, and manage status pages the same way you would, through a scoped API token you control. Monitoring becomes something your automation does, not another dashboard to babysit.
# your agent, over the Undownable MCP server
agent → create-monitor
{ "type": "http", "url": "https://api.acme.dev",
"interval": 60, "name": "acme api" }
✅ monitor "acme api" created, checking every 60s
# minutes later, when it dips
agent → list-incidents { "status": "open" }
🔴 acme api: HTTP 500, open for 00:02:14
Self-hosted monitoring shares fate with the network it watches: when your lab, your ISP, or the monitor itself goes down, nobody gets paged. Undownable watches from the outside, so the one failure your stack can't self-report still raises an alert.
# your side (cron, Kuma push, systemd timer…)
*/5 * * * * curl -fsS \
https://undownable.com/ping/01JZ…X4
# our side, when the pings stop
🔴 homelab-kuma is down: missed heartbeat
✅ homelab-kuma recovered after 00:12:41
One flat price per team: no per-seat, per-monitor, or per-status-page charges. Invite your whole team at no extra cost, with every alert channel on every plan.
$0/mo
$8/mo
billed yearly at $100, or $10 month-to-month
Starts with a 14-day free trial. No card required.