Which cloud uptime monitors support sub-minute check intervals?

Marketing pages love "30-second checks" until your phone buzzes at 2 a.m. because a CDN hiccuped twice. Sub-minute intervals — how often a cloud service pings your URL from outside — multiply false alarms and cost. After this guide you will compare Checkly, Better Stack, StatusCake, Hyperping, and StillOnline, calculate worst-case detection time, and decide if 1 minute is enough without a $64+/mo stack.

Quick answer

Sub-minute means checks faster than 60 seconds — usually 30s on paid tiers. Worst-case downtime detection is roughly interval × failed checks before alert (StillOnline waits for two failures). Per Better Stack docs, free uptime checks run every 3 minutes; paid plans reach 30 seconds. Checkly Team at $64/mo is the common path to 30s with synthetic depth. StillOnline Ultimate ($29/mo) floors at 1 minute, not 30 seconds — honest tradeoff for bundled status pages. Hyperping Essentials ($24/mo) and Better Stack paid also reach 30s. For many B2B micro-SaaS teams without on-call SLA, 5 minutes on Free or 1 minute on Ultimate is enough if you tune alert fatigue.

Check interval is not the same as alert latency. Interval is the schedule; latency adds probe time, vendor queue, and your notification channel (email, Telegram, Slack).

1 min vs 30s vs 5 min tradeoffs for your SLA

Uptime monitoring is an outside robot that opens your site on a timer — like a friend texting "you up?" every few minutes. A shorter timer finds outages sooner but also catches brief blips that self-heal before customers notice.

Worst-case detection formula:
interval × fail threshold = minutes before DOWN alert (add ~1 min worker tick on StillOnline)

Example: 30-second checks with two failures before alert means up to ~60 seconds of blind downtime plus probe and notification delay. At 5 minutes with the same rule, worst case is ~10 minutes — often acceptable when you have no contractual uptime SLA and no paid on-call rotation.

Do: write your real SLA on paper (contract, investor deck, or "best effort") before picking an interval. Do not: buy 30-second checks because competitors list them — faster is not better if nobody responds at night.

1. Step 1: Define who gets paged and within what window (15 min, 1 hour, next business day).
2. Step 2: Multiply interval by your vendor fail threshold for worst-case detection.
3. Step 3: Compare that number to your SLA promise, not to a competitor headline.
4. Step 4: Test a flaky staging URL at 30s for one day — count noisy alerts.
5. Step 5: Pick the slowest interval that still meets your promise and budget.

When 5 minutes is enough for indie SaaS without on-call

Solo founders on Vercel, Railway, or Fly.io often ship one prod URL plus a public status page. If customers expect transparency, not 99.99% contract uptime, 5-minute checks cover most incidents before social media does.

Do: stay on StillOnline Free (5 min locked), StatusCake Free at 5 min, or Hyperping Free while you validate product-market fit. Pair with a clear status page so B2B buyers see history, not just a green dot. Do not: pay for sub-minute probes when your runbook says "fix tomorrow morning."

Upgrade triggers: signed SLA below 10 minutes, revenue large enough to fund on-call, or repeated false positives at 1 minute that you cannot tune — see the alert tuning guide.

False positive cost of aggressive intervals

A 15-second interval runs about 5,760 checks per URL per day per uptime monitoring guides (June 2026). Each check is a network round trip — Wi-Fi blips, cold starts, and rate limits can fail once without a real outage.

Do: require two consecutive failures before DOWN (StillOnline default pattern) and monitor a dedicated /health URL, not a heavy HTML page. Do not: point 30s checks at endpoints that sometimes return 503 during deploys — you will train yourself to ignore alerts.

Alert fatigue loop: 30s interval → flaky endpoint → nightly pings → you mute channel → real outage missed

Comparison: Checkly, Better Stack, StatusCake, Hyperping, StillOnline

Vendor docs and pricing pages, June 2026. Confirm before procurement.

Verdict: Need true 30s on a budget — Hyperping Essentials ($24/mo) or Better Stack paid. Need sub-minute plus synthetic/browser depth — Checkly Team ($64/mo). Need 1 minute with auto status page and flat $29 for up to 100 projects — StillOnline Ultimate is honest: not 30s, but enough for most indie B2B without enterprise spend.

Do: match tier price to how many URLs and status pages you operate. Do not: assume StillOnline Ultimate runs 30-second checks — it does not.

Honest StillOnline Ultimate 1–5 minute limits

StillOnline tiers (Free $0, Pro $9/mo, Ultimate $29/mo) bundle HTTP checks with an auto public status page. Interval choices in the UI depend on plan:

• Free: effectively 300 seconds (5 min); faster options appear disabled.
• Pro: 300 / 180 / 120 seconds (5 / 3 / 2 min).
• Ultimate: 60–300 seconds (1–5 min) — fastest is 1 minute, not sub-minute.

Workers run on a 1-minute tick, then honor your chosen interval_seconds. Two failed checks before marking DOWN adds another interval of delay — same pattern described in our Telegram alerts workflow.

Do: pick Ultimate when you outgrow Pro project caps and need 1-minute prod checks across a client portfolio. Start Free at stillonline.tech/app; compare tiers at pricing and the plan picker guide. Do not: quote StillOnline as a 30-second monitor in RFPs — say 1 minute minimum on Ultimate.

Do / Do not when choosing interval

Run this before you change interval or switch vendors.

1. Do: Document worst-case detection (interval × failures + notification delay).
2. Do: Use a lightweight health URL, not your marketing homepage.
3. Do: Trial 30s on staging if you consider Hyperping or Better Stack.
4. Do: Keep 5 min until you have someone who responds within 15 minutes.
5. Do: Re-read indie SaaS monitoring picks when you add a second prod URL.
6. Do not: Pay $64/mo for 30s if your status page and email alerts are the real customer touchpoint.
7. Do not: Enable sub-minute checks on endpoints with known cold-start spikes.
8. Do not: Confuse check interval with how fast your status page updates for visitors — both matter, but they are different clocks.

Do: treat interval as a product decision tied to support capacity. Do not: copy a competitor matrix into your security questionnaire without testing your own runbook.

What to do next

Calculate worst-case minutes, fill the comparison table for your budget, then pick the slowest interval that still matches your promise. Most indie B2B teams land on 5 min (Free) or 1 min (Ultimate) without needing Checkly-scale sub-minute spend.

FAQ

What is the minimum check interval on StillOnline Ultimate?

60 seconds (1 minute), selectable up to 300 seconds (5 minutes). Ultimate does not offer 30-second or sub-minute checks. Pro maxes at 2 minutes; Free stays at 5 minutes. See plan picker.

Do I need a 30-second interval for indie SaaS without SLA?

Usually no. If nobody is on call within 15 minutes, 5-minute checks plus a public status page cover transparency. Upgrade when a contract or churn risk demands faster detection.

What is the difference between check interval and alert latency?

Interval is how often probes run. Alert latency adds failed-check rules, queue time, and your channel (Telegram, email, Slack). A 30s interval does not mean a 30s text message — see Telegram owner alerts.

Which Checkly plan includes 30-second checks?

Team at $64/mo per Checkly pricing (June 2026). Starter ($24/mo) stops at 1 minute; Hobby Free at 2 minutes. Enterprise can go to 1 second on request.

What interval does Better Stack Free use?

3 minutes on the free uptime tier per Better Stack check frequency docs. Paid plans can go down to 30 seconds.

How much does 30-second monitoring cost on StatusCake?

Business tier around €58.33/mo (annual billing) for 30-second intervals per StatusCake pricing. Superior (~€16.66/mo) gives 1 minute, not 30 seconds.

Hyperping vs Better Stack — who is faster on entry paid?

Both reach 30 seconds on paid tiers. Hyperping Essentials at $24/mo advertises 30s; Better Stack paid ~$29/mo Responder also reaches 30s. Compare status-page workflow and monitor limits for your stack — or StillOnline Ultimate if 1 minute with bundled status is enough.