Ceiling, not floor
Average spacing might be 44.6 minutes. Your scheduler usually needs an integer minute or second step. Rounding down makes you poll too often and risk blowing the quota.
Cron spacing for monthly API quotas
API Rate Limit & Cron Frequency Calculator
Spread a fixed monthly API budget across days and minutes: safe spacing between calls, month-length comparison, and cron or systemd guidance so free-tier limits never surprise you.
By Jeff Beem
Technical spec
Rate limits
01
Primary quota
Optional burst cap
Many APIs enforce both a monthly quota and a short-term burst cap (e.g. 10 req/min). If both apply, the calculator uses whichever constraint produces the longer safe interval (the bottleneck).
02
Planning window (month)
03
Safety margin
Increase this if your client auto-retries on failure, you run workers in multiple regions, or you cannot centrally coordinate concurrent callers. 5% is a safe default for single-region, single-process polling.
Output
Throttle & schedule
A
Bottleneck engine
| Constraint | Base interval | After margin |
|---|---|---|
| 1,000 per month (strictest) | 44.64 min (2678 sec) | 46.87 min (2812 sec) |
Safe interval (bottleneck)
46.87 min (2812 sec)
Limited by 1,000 per month after 5% margin.
The address bar updates as you edit, use share to copy the full link.
Monthly budget (1,000 / month cap)
Safe firings use floor(monthSeconds ÷ interval)
February (common)860 / 1,000 (140 left)
February (leap)890 / 1,000 (110 left)
30-day months921 / 1,000 (79 left)
31-day months952 / 1,000 (48 left)
B
Cron & systemd
Cron minute field
*/N * * * * only equals “every N minutes” when N is one of 1, 2, 3, 4, 5, 6, 10, 12, 15, 20, 30 (divisors of 60). Otherwise use systemd or staggered lines (see 45‑minute template).
Suggested multi-line cron
# Illustrative only, cron cannot hit 2812.32s (~46.87 min) exactly; prefer the systemd timer below. # Example: split across three offsets in the hour (adjust path and verify times): 0 * * * * /path/to/your/script.sh 20 * * * * /path/to/your/script.sh 40 * * * * /path/to/your/script.sh
systemd timer (accurate interval)
[Timer] OnUnitActiveSec=2812.32s AccuracySec=1s
No single crontab minute pattern matches this spacing exactly. Prefer the systemd timer below; the multi-line block is illustrative.
C
Code snippets
import time
INTERVAL_SEC = 2812.32
while True:
# your_api_call()
time.sleep(INTERVAL_SEC)
Common API Limits
Illustrative limits, confirm on each provider’s site. “Approx. spacing” opens this page with that quota pre-filled (shareable URL state).
| API | Quota | Window | Approx. spacing | Docs |
|---|---|---|---|---|
| Abstract API | 1,000 | Monthly | ~45 min | Plans & limits |
| Discord API | 10,000 | Per minute | ~6 ms | Rate limits |
| GitHub API (authenticated) | 5,000 | Per hour | ~0.8 s | Rate limits |
| ipinfo.io | 50,000 | Monthly | ~1 min | Plans |
| NewsAPI | 100 | Daily | ~14.4 min | Pricing |
| OpenAI API (free) | 60 | Per minute | ~1 s | Rate limits |
| OpenWeatherMap | 1,000 | Monthly | ~45 min | Pricing |
| Reddit API | 60 | Per minute | ~1 s | API wiki |
| Shopify API (basic) | 120 | Per minute | ~0.5 s | Rate limits |
| Slack API (tier 3) | 50 | Per minute | ~1.3 s | Rate limits |
| Stripe API (live) | 6,000 | Per minute | ~10 ms | Rate limits |
| Twitter / X API (free) | 50 | Per minute | ~1.3 s | Rate limits |
Monthly API quotas and steady polling
Free tiers often quote a maximum number of HTTP calls per month. If you poll on a timer, the safe period is driven by calendar length and integer rounding, not only “divide by 30.” Here is how to reason about spacing, cron, and drift.
Ideas worth keeping straight
Cron is not “every N minutes”
Minute fields reset each hour. Patterns like */45 do not mean one fire every 45 minutes across hour boundaries. Read the tool warning before you paste a line into crontab.
Compare month lengths
The bar comparison shows how many firings fit in 28-, 29-, 30-, and 31-day months at your chosen spacing. If you planned for February only, March might overshoot.
Retries cost quota
401/429 responses and client retries still hit the API. Leave slack under the cap if your client automatically retries.
API Rate Limit & Cron Frequency Calculator
Translate monthly API quotas into safe polling intervals. Plan cron and systemd timers without exceeding 1k, 5k, or 10k monthly calls.
What This Tool Does
You enter how many HTTP requests your plan allows per month and how many days you want to bake into spacing (default 31). The calculator outputs average minutes between calls, a rounded-up recommendation, and how many firings fit in 28-, 29-, 30-, and 31-day months at that spacing. It flags when naive cron syntax would overfire and suggests safer options.
How to Calculate API Frequency for Monthly Quotas
Let be your monthly maximum and the number of days you treat as one billing month for planning. Total minutes in that window is . Average minutes between requests is . A practical scheduler uses a whole number of minutes, , so rounding never speeds you up into an overage.
Why Plan Around the Longest Month You Care About
Providers usually count calendar months. A schedule tuned only for February leaves extra slack in March, but a schedule tuned only for a 30-day month can overshoot when the month has 31 days. Choosing up front is an easy hedge unless you intentionally rotate keys or change cadence mid-year.
Cron and Timers Without Breaking the Limit
When your minute spacing divides 60 evenly, a cron line like can express “every N minutes.” Otherwise use a systemd timer (for example OnUnitActiveSec=Nmin), Kubernetes interval schedules, CI cron with explicit RFC3339, or a short shell loop with sleep. For exactly 45-minute spacing, three staggered cron lines are the usual portable Unix pattern; the calculator surfaces them when relevant.
Quick reference: common quotas on a 31-day basis
Illustrative values using minutes: 1,000 calls needs at least 45 minutes between calls; 5,000 calls needs at least 9 minutes; 10,000 calls needs at least 5 minutes. Your provider’s exact window (rolling 30 days vs calendar month) may differ, so treat these as starting points.
API Request Frequency Calculator FAQ
How many minutes between requests for 1,000 API calls per month?
Plan using the longest month you care about. For 31 days there are minutes. Dividing by 1,000 gives about 44.64 minutes per request on average. Rounding up to 45 whole minutes keeps you under 1,000 calls in that month after rounding and small clock drift. The calculator does this ceiling step for any quota you enter.
Why should I base spacing on 31 days?
If you size your job for a 28- or 30-day month, a 31-day month has extra minutes and extra run opportunities. Fixed cron schedules often fire the same number of times per day, so you can creep over a strict monthly maximum. Using 31 days (or the calculator default) trades a slightly slower cadence for staying under the cap every month.
Can I use */45 in cron for every 45 minutes?
No. In standard cron, runs at minute 0 and 45 of every hour, which is about 48 times per day, not one call every 45 minutes. For a true 45-minute cadence you need multiple cron lines that stagger across hours, a systemd timer with OnUnitActiveSec=45min, Kubernetes CronJob with interval semantics, or a loop script with sleep. The tool shows a three-line cron template when your safe spacing lands on 45 minutes.
When does a simple */N * * * * cron line work?
When N minutes divides 60 evenly (for example 5, 10, 15, 20, 30), fires exactly every N minutes within the hour and lines up across hours. For other intervals, use a scheduler that supports wall-clock intervals, not minute-star tricks alone.
How accurate is “requests used” for shorter months?
The table assumes one request exactly every N minutes for the whole month ( ). Manual retries, failures, or bursts still count against your provider. Treat the numbers as a schedule baseline, not a legal guarantee.
How often can I call the OpenWeatherMap free tier?
The common free tier is often around 1,000 calls per calendar month. Spread evenly across a long month with a little buffer, that implies on the order of one call every ~45 minutes, use the calculator for your exact quota. Open pre-filled for 1,000/month.
What is the difference between a calendar month and a rolling window API limit?
A calendar month resets on the 1st (or your provider’s billing anchor) and uses that month’s actual length (28–31 days). A rolling window looks back exactly a fixed duration (often 30 days from “now”), so the effective budget does not swell in 31‑day months. Stripe, OpenAI, and many SaaS APIs describe rolling limits in their docs.
My cron */45 job isn’t firing every 45 minutes, why?
In standard cron,
*/45 runs at minute 0 and 45 of each hour (~32 times/day), not one execution every 45 minutes wall-clock across hour boundaries. Use multiple staggered cron lines (this tool prints a template when relevant), or a systemd timer / sleep loop, see the amber “Cron minute field” note above the systemd block.What happens if I have both a monthly limit and a per-minute burst cap?
Enter the monthly quota as your primary constraint, then use + Add burst limit for the short-term cap (e.g. requests per minute). The bottleneck table shows which rule produces the longest safe interval; that spacing keeps you within both limits when you throttle to the strictest constraint.
Mathematical Reference Note
Calculation Logic: This tool uses standard mathematical algorithms. While we strive for accuracy, errors in logic or user input can result in incorrect data.
Verification: Results should be cross-checked if used for important academic, professional, or personal calculations.
Standard Terms: This tool is provided free of charge and as-is. CalcRegistry provides no warranty regarding the accuracy or fitness of these results for your specific needs.
© 2026 CalcRegistry Reference Last System Check: May 2026Free Online Utility Tools