Everything you love about simple monitoring, with more reliability, more integrations, and no self-hosting.
| Name | Type | Status | Uptime (30d) |
|---|---|---|---|
| Website | HTTP | Up | 99.99% |
| API | HTTP | Up | 99.98% |
| Database | TCP | Up | 99.95% |
| DNS | DNS | Up | 99.99% |
| Cron job | Heartbeat | Up | 99.97% |
It isn't a knock on Kuma — it's true of anything that monitors from inside your own failure domain, ours included. If the hypervisor dies, the UPS runs out, or your ISP drops at 3am, the monitor that would have told you is part of the outage. A watchdog can't live on the thing it watches. We wrote up the full argument here.
Point a PylonMon HTTP check at your Kuma instance and a heartbeat at the host it runs on. Kuma keeps watching your LAN in detail; we page you when Kuma itself — or the box, or the whole site — goes quiet. The free tier covers this completely.
HTTP/S, TCP, ping, DNS, TLS-cert and heartbeat monitors from our infrastructure, alerts to Discord, Slack, Telegram, ntfy, email, webhooks — and SMS & voice calls on paid tiers. No container to update, no SQLite to back up, no reverse proxy to expose.
A tiny open-source agent (MIT) reports CPU, memory and per-mount disk — and if a machine goes quiet, that silence is itself the alert. Three servers free.
| Uptime Kuma | PylonMon | |
|---|---|---|
| Runs on | Your hardware, inside your network | Our infrastructure, outside it |
| Survives your outage | No — it IS part of the outage | Yes — that's the point |
| Cost | Free, unlimited monitors | Free for 20 monitors; paid from $9/mo |
| Maintenance | Updates, backups, the host itself | None — it's hosted |
| LAN-internal checks | Excellent — it lives there | Via the agent & heartbeats |
| SMS / voice-call paging | Via third-party services you wire up | Built in on paid tiers |
| Probing private LAN IPs | Yes — it lives there | Local probes via the beacon agent |
If self-hosting everything is the goal, keep Kuma — sincerely. This page exists for the day you realize nobody is watching the watcher.
Not by polling — we're outside your network by design, and a private IP is unreachable from the internet. Instead: put a heartbeat on internal jobs (they check in outward), or the beacon agent on internal machines (it pushes vitals out). Silence from either pages you.
No — the pairing works well: Kuma for LAN detail, PylonMon as the off-site dead-man's-switch over Kuma, its host, and everything public. But you don't need both to be covered — between hosted monitors, heartbeats, the agent, and local probes via the beacon, PylonMon can fully replace it.
Free forever for 20 monitors, 1-minute intervals, heartbeat + HTTP/S + TCP + ping, and 3 agent-monitored servers. Paid tiers start at $9/mo for 30-second intervals, more monitors, SMS & voice, and extra monitor types. See pricing →
Free for 20 monitors. Two-minute setup. Kuma can stay.
