Flat-rate. Unlimited team.
No surprises.

One endpoint we check — an HTTP/HTTPS URL, a TCP port, a DNS record, or an SSL certificate. Each check counts as one monitor. We don't charge separately for the number of regions we check from or the frequency.

Yes. 25 monitors, 3-minute checks, commercial use OK, no card required, forever. It says so in the terms, not just here. Free checks run every 3 minutes; UptimeRobot's free plan checks every 5. Free stays free; 1-minute detection starts at $10.

A soft cap with an in-app prompt to upgrade — never a surprise overage bill. Your existing monitors keep working; adding one above the cap is blocked with a message naming the plan that lifts it. We'd rather lose a few dollars of overage revenue than wake you up to an unexpected $400 invoice.

No. Never. Every paid plan has unlimited team members — add your whole engineering org, your SRE rotation, your CEO who wants alerts, same price. Free includes 3 seats, which covers a side project honestly; per-seat pricing is one of the biggest reasons people hate enterprise tools, and no Repose plan will ever have it.

Every paged incident can get an AI root-cause analysis — category, likely cause, user impact, what to check first — built from what the probes actually saw. Free workspaces get 10 investigations a month; paid plans are unlimited.

Asking Mercury follow-up questions on an incident page is metered separately, because that one has no natural ceiling: 30 questions a month on Free, 100 on Indie, 400 on Team, 1,500 on Business, 5,000 on Scale. Those are set well above what debugging real outages costs — they're there so a runaway script can't spend more than your subscription, not to ration the feature. If you hit one while genuinely working, email us and we'll raise it.

The deterministic parts — evidence capture, the audit trail, consensus — are never capped on any plan.

Because it's the feature that makes an incident page tell you something a probe can't. Probes show symptoms — a 503, five regions, a timestamp. A deploy marker turns that into “it started 1m 45s after a release”, which is usually where you'd start looking anyway.

One thing worth saying plainly: we record your deploys on every plan, including Free. Only seeing them on the incident page is gated. Correlation is worth something because of history, and a feature that started collecting the day you upgraded would be useless for a quarter. Wire it up now on Free and the record is already there whenever you want to look at it.

Not underneath you. The terms of the plan you're on never get worse while you're on it — if we ever change limits or pricing, changes apply to new signups, and existing customers keep what they signed up for. Changing the deal on people who already built on it burns trust that takes years to earn back, and trust is our entire product.

Three things that matter on day one: (1) multi-region consensus alerting on by default — single-region flaps die in the dashboard, not on your phone; (2) dependency-aware alerts that tag upstream incidents from Stripe, AWS, Cloudflare, Vercel etc. before paging you; (3) AI root-cause analysis on every plan — 10 investigations a month free, unlimited on paid. UptimeRobot has the widest free tier in the category; BetterStack bundles a lot and is growing into a full observability platform. We stay focused on uptime, on-call and status pages, and on not paging you for nothing.

Yes. Copy your list of URLs — or paste the whole CSV export — into Import a list in the app. It reads one monitor per line, shows you exactly what it will create, and makes them all at once as HTTP checks. It doesn't connect to the other tool's API and it doesn't copy alert rules or history; keyword, port and heartbeat checks are quick to add by hand afterwards. Keep the old tool running for a fortnight and compare what each one woke you for.

99.9% on Team, 99.95% on Business, 99.99% on Scale. We publish our own status page and the live health of the probe fleet at repose.sh/probes. When we go down, you'll know — credits are automatic on the affected tiers.

Yes to both. Up or down, monthly or annual, prorated to the day. Cancel from the billing portal in Settings — no "talk to retention" calls, no friction. Annual customers who cancel get a prorated refund for the unused months.