10pm uptime vs UptimeRobot vs Uptime Kuma
Three different answers to "where does the check actually run from". UptimeRobot probes from its own cloud locations; Uptime Kuma is a single binary you host and run yourself; 10pm uptime does both — checks from our own fleet out of the box, plus an optional open-source agent you run anywhere, all reporting into a control plane we operate.
| Feature | 10pm uptime | UptimeRobot | Uptime Kuma |
|---|---|---|---|
| Where checks run | Our fleet of monitors worldwide — and, optionally, your own hardware via the open-source agent | Their cloud locations only | Wherever you host it yourself |
| Reaches internal / firewalled services | Yes — run the optional agent inside your network. Public targets need no agent at all | No built-in agent; heartbeat/cron workaround only | Yes, if you host the whole thing there |
| Hosting & upkeep | None — the dashboard and alerting are managed for you. The agent is only needed for internal targets | None | Yours: updates, backups, and keeping the monitor itself up |
| Multi-organisation / team accounts | Built in | Team & Scale plans only | No — one admin login, no per-user access control |
| Check types | http(s), TCP, bare TLS handshake, heartbeat | HTTP, keyword, ping, port, DNS, SSL, heartbeat | http(s), TCP, DNS, ping, push, Docker, Steam |
| Knowing when your monitoring itself goes quiet e.g. an internal Uptime Kuma, a cron job, a backup script |
Heartbeat checks — point anything that can call a URL at us; we alert when it stops. Works with a plain Kuma HTTP(s) monitor | Heartbeat monitor | Push monitors receive beats at Kuma's own URL — nothing outside your network notices if Kuma itself dies |
| Cert-expiry warnings on non-HTTP services e.g. SMTPS, IMAPS — anything that's TLS with no HTTP on top |
Yes | No — SSL monitor is HTTPS-only | No — tied to an HTTP(S) monitor |
| Alert channels | Push (ntfy), email, webhook and Slack, routed per check | Email, SMS, voice call, app push, integrations | 90+ services via Apprise; no SMS or voice |
| Public status page | Yes, on paid plans | Yes | Yes |
| Source availability | Agent is open source | Closed | Fully open source |
| Pricing | Early access — see plans | Free tier, paid from ~$13/mo | Free — your own infrastructure cost only |
Reflects each project's publicly documented capabilities as of September 2026. UptimeRobot and Uptime Kuma are trademarks of their respective owners; we're not affiliated with either. Their feature sets change — check their own sites for the current word.
Picking UptimeRobot?
A great fit if everything you're checking is already public on the internet. The moment one target sits behind a firewall — an internal admin panel, a VPN-only service, a branch office link — there's no first-party way to reach it.
Picking Uptime Kuma?
The right call if you want to run and own every piece yourself. 10pm uptime keeps Kuma's "the probe is just a binary I run" idea, but hands the dashboard, alert routing, and account management to us instead of one more service on your box. Already running Kuma? Keep it, and point a heartbeat at us so you hear about it when Kuma itself goes down.
Where 10pm uptime fits
Start with our shared fleet for anything public, add your own agent(s) only for what sits behind a firewall, and let a managed control plane handle everything that isn't the probe itself — without standing up a second instance.