#1★ TOP PICK
Uptime Kuma
The self-hosted monitor that beat the paid ones on features.
96
OPEN SOURCEMITSELF-HOSTLOCAL-FIRST
Uptime Kuma is a self-hosted monitoring tool that does, for free and on your own server, most of what the paid services charge monthly for: HTTP, TCP, ping, DNS and keyword checks, 60-second intervals, a genuinely good dashboard, public status pages, and notifications to more than ninety services including Slack, Discord, Telegram, email and webhooks. It runs as one Docker container on hardware you already have. At roughly 89.7k GitHub stars it is one of the most-starred self-hosted projects in existence, and that popularity is deserved rather than fashionable — the free tiers of the commercial services are more limited than this is. MIT licensed, pushed to daily.
Strengths
- +60-second checks and unlimited monitors with no plan to upgrade to
- +Public status pages included rather than sold separately
- +Notifications to 90+ services out of the box
- +MIT, one container, and the most-starred tool in the category
Trade-offs
- −It runs on your infrastructure — so it cannot tell you your infrastructure is down
- −Single-node by design; no multi-region checking
- −You own updates, backups and the uptime of the uptime monitor
- −No phone-call escalation or on-call rotation
Free and open source. One Docker container on a server you already run. #2
Gatus
Monitoring as a YAML file, for people who already run everything from Git.
94
OPEN SOURCEApache-2.0SELF-HOSTLOCAL-FIRST
Gatus is a developer-first health-check service configured entirely in YAML, which means your monitoring lives in the repository, gets reviewed in pull requests, and is deployed like any other code. It handles HTTP, TCP, ICMP, DNS and TLS-expiry checks, evaluates rich conditions on responses — status code, body content, response time, certificate age — and ships a clean dashboard and status page. Apache-2.0, around 11.7k stars, actively developed. It is the natural pick for a team already doing GitOps, and the wrong one if a non-engineer needs to add a monitor.
Strengths
- +Configuration is YAML in Git — reviewed, versioned and diffable
- +Rich condition evaluation, including response body and TLS expiry
- +Apache-2.0 with no enterprise carve-out
- +Very light: one small binary, minimal memory
Trade-offs
- −No UI for adding monitors — everything is a config change
- −Same blind spot as any self-hosted monitor: it can go down with you
- −Fewer notification integrations than Uptime Kuma
- −Not usable by non-technical team members
Free and open source. A single binary or container. #3
Healthchecks.io
Monitors the jobs that fail silently — the cron nobody noticed stopped.
93
OPEN SOURCEBSD-3-ClauseSELF-HOSTLOCAL-FIRST
Healthchecks solves the opposite problem from the rest of this page. Instead of pinging your site, it waits for your scheduled jobs to ping it, and alerts you when one does not — the nightly backup that quietly stopped running three weeks ago, the certificate renewal that failed, the data export nobody checked. That class of failure is invisible to every uptime monitor here, and it is usually the more expensive one. BSD-3-Clause, around 10.2k stars, self-hostable in Python with a hosted tier that has a free plan. It complements an uptime monitor rather than replacing one.
Strengths
- +Catches silent cron and backup failures nothing else on this page sees
- +Dead simple integration — one curl at the end of your job
- +BSD-3-Clause, self-hostable, with an honest hosted free tier
- +Notifications to email, Slack, PagerDuty, webhooks and more
Trade-offs
- −Not an uptime monitor — it does not check whether your site responds
- −Only sees jobs you have instrumented to report in
- −Self-hosting means running a Django application and a database
- −Complements rather than replaces the rest of this comparison
Free and open source to self-host. Hosted plans have a free tier for a small number of checks. #4
UptimeRobot
Hosted checks from outside your network, with a free tier that is actually usable.
48
SOURCE-AVAILABLEProprietary (hosted service)
UptimeRobot is the hosted answer, and it solves the one problem no self-hosted monitor can: it watches your site from outside your own infrastructure, so when your server, your network or your data centre goes down, the thing telling you about it is not also down. The free tier is unusually generous — 50 monitors at 5-minute intervals, email alerts and a public status page — which is more than enough for most personal sites and small businesses. Paid plans add 30-second checks, SMS and voice alerts, and multi-location verification to cut false alarms. It is proprietary and hosted, which is why it scores where it does; it is on this page because external monitoring is a real requirement and pretending otherwise would not serve the reader.
Strengths
- +Checks from outside your network — the one thing self-hosting cannot do
- +Free tier covers 50 monitors, which is genuinely enough for most people
- +Nothing to install, patch or keep running
- +Multi-location verification cuts false alarms from transient network blips
Trade-offs
- −Proprietary and hosted — you do not own the monitoring or its history
- −Free tier is capped at 5-minute checks, so a short outage can be missed
- −SMS and voice alerts are paid add-ons
- −Your uptime data lives in a third party's system
Free tier: 50 monitors at 5-minute checks, with a public status page. Paid plans add 30-second intervals, SMS/voice alerts and multi-location checks. Checked 2026-08-01. #5
Better Stack
Hosted monitoring with real on-call — when someone has to be woken up.
40
SOURCE-AVAILABLEProprietary (hosted service)
Better Stack combines uptime monitoring, incident management, on-call scheduling and log management in one hosted product. It is the option to consider when the requirement is not really monitoring but escalation: rotations, phone calls that keep ringing, acknowledgement tracking and post-incident timelines. None of the open tools here do that, and building it is not a weekend project. Proprietary and hosted with a free tier. Listed because for a team with a genuine on-call obligation it is a different category of answer, and pretending Uptime Kuma covers it would be misleading.
Strengths
- +Real on-call rotations and phone escalation, which nothing open here offers
- +Incident management and status pages in the same product
- +30-second checks and multi-region verification
- +Modern interface that teams actually adopt
Trade-offs
- −Proprietary and hosted — no self-host path at any price
- −Per-seat on-call pricing grows with the team
- −More product than most people needing an uptime check require
- −Your incident history lives in their system
Free tier available; paid plans per monitor and per on-call seat. Checked 2026-08-01.