Calculators That Run in Your Browser

Everyday sums, worked properly

Most of these tools answer a question you could do by hand but would rather not get wrong. The Tip & Bill Splitter settles a restaurant bill. The Meeting Cost Calculator turns headcount and hourly rates into the price of an hour. The Roasting Time Calculator gives a cooking time from a joint's weight. Small jobs, cleared in seconds.

Others carry more weight. The Retirement Calculator, Savings Goal Calculator, Mortgage Affordability Calculator and Net Worth Calculator all deal with money over years. Compounding makes that kind of mental arithmetic unreliable. The Sample Size Calculator and Big Number Calculator are work tools. The Screen Size Calculator, Download Time Calculator, Time Zone Converter and Event Budget Planner sit between the two: quick, but tedious to repeat by hand.

What a browser is good at here, and what it is not

Arithmetic is the one job a browser already does natively. There is nothing to install and nothing to convert, so these pages are small, open quickly and keep working with the network off. Nothing you type leaves the machine, which matters when the inputs are your salary, your savings or what you owe.

The real constraint is not speed but assumptions. A retirement or affordability figure is only as good as the growth rate, inflation and interest rate you feed it. A browser cannot know any of those. Read the result as a shape rather than a promise, and take proper advice before acting on it.

Two smaller limits are worth knowing. Ordinary computer numbers stop being exact past roughly fifteen digits, so very large or very precise work needs a tool built for that. And date rules change by government decision, so date work is only as current as the rules your browser carries.

How to pick, and the mistake to avoid

Choose by the unknown, not the topic. The Savings Goal Calculator answers how much to put aside each month. The Retirement Calculator answers whether a pot will last. Same family of maths, different missing number. Naming your unknown first usually picks the tool in one go.

The common mistake is trusting a single run. Where a tool asks for a rate you are guessing at, try a pessimistic value and an optimistic one. See how far apart the answers land. One percentage point can move a thirty-year figure a long way. Then check units before you believe anything: monthly against yearly, gross against net. Most wrong answers here come from the input, not the sum.

Date sums, and the one-day error hiding in them

Every tool here that takes a date reads it through the same small piece of shared code, and that code exists because of one bug. Hand the text 2026-07-29 straight to a browser and it is read as UTC, which lands on 28 July for anyone west of Greenwich. The shared parser pulls the year, month and day apart and builds a local date instead, so a birthday typed in Chicago is the birthday and not the day before it.

Two more corrections sit next to it. Day counts move both ends of the span to midday before subtracting, so the hour that disappears when the clocks go forward cannot shave a whole day off the answer. And the age-shaped result — two years, three months, five days — borrows from the real length of the month before, not a flat thirty. That is how a person counts and it is not what plain subtraction gives you. The Date & Age Calculator and Days Between Dates both run on it.

Time zones are the one part the browser already does better than we could. The Time Zone Converter asks Intl for the browser's own IANA zone database, so daylight saving comes out right for every zone and every date, including the weeks when one country has switched over and the other has not. The Working Days Calculator goes the opposite way and ships no holiday list at all. You set which days count as your working week — Sunday to Thursday is normal across much of the Middle East — and paste your own public holidays into a box. A built-in calendar would be wrong for most visitors and stale for the rest.

Where two formulas disagree, you get both

Several tools here refuse on purpose to hand back a single number. The Ideal Weight Calculator prints four: Devine (1974), Robinson (1983), Miller (1983) and Hamwi (1964). All four have the same shape, a base weight plus a step for every inch over five feet, and they disagree because they were written for different jobs — Devine was written for working out medicine doses. Printing one of them would suggest a precision none of them has, so the spread is the answer.

The One-Rep Max Calculator runs Epley and Brzycki together and headlines the average of the two. They agree closely up to about ten reps and pull apart after that, and the tool will not accept a set longer than fifteen reps at all rather than return a figure it does not believe. The Blood Pressure Checker shows the 2017 ACC/AHA thresholds beside the older 140/90 line, because one of them calls 130/80 hypertensive and the other does not, and which one applies depends on where you are and who is treating you. It also grades a reading by whichever of the two numbers is worse, so 118/95 is not waved through because the top figure looks normal.

The Distance & Bearing Calculator turns the disagreement into the whole point. Haversine treats the Earth as a sphere and is out by up to roughly half a percent, about five kilometres on a thousand-kilometre leg. Vincenty solves the same problem on the WGS-84 ellipsoid that GPS itself uses and is right to a fraction of a millimetre, at the cost of an iteration that can fail to settle for points nearly opposite each other on the globe. When that happens the tool says so and falls back, instead of quietly handing you a wrong number. It reports two bearings for the same reason: a great circle from London to New York sets off at about 288° and arrives on about 237°.

What the page keeps, and what you can carry away from it

One hundred and four of the hundred and twenty tools share the same shell, and it does three things worth knowing about. It recalculates as you type, after a pause of about 140 milliseconds so the figures do not flicker on every keystroke. It writes your inputs into the browser's own storage under a key belonging to that one tool, so the numbers are still sitting there tomorrow — on that machine, in that browser, and nowhere else. The reset button on the page deletes that entry.

Currency is guessed rather than assumed. The shell reads the region out of your browser's language setting and picks from fourteen: the Indian rupee, US dollar, euro, pound, Australian and Canadian dollars, dirham, Singapore dollar, yen, rand, naira, Pakistani rupee, taka and Brazilian real, falling back to the dollar when it cannot tell. It also shortens big amounts the way a local reader expects them — lakh and crore for the Indian, Pakistani and Bangladeshi currencies, thousands, millions and billions for the rest.

Every result carries a save bar with three buttons: a plain .txt file, a PNG image, and copy to the clipboard. The saved copy always carries the inputs that produced it, read back off the page with their real labels, so a dropdown records "Reducing balance" rather than the number one. A figure with no record of what produced it is not much use six months later. Where a table runs long the image trims it and prints how many rows it left out, while the .txt file keeps all of them. Every one of those files is written by your own browser — nothing is sent anywhere at any point.