How We Round & Format Numbers
Every result on this site follows the same, simple rules — explained here in full, so you never have to guess whether a number has been rounded, truncated, or shown exactly.
The rule: exact numbers in full, everything else to 6 significant figures
Two cases, and only two:
1. If a conversion works out to an exact whole number — like 1 day = 24 hours, or 1 pebibyte = exactly 1,125,899,906,842,624 bytes — it's shown in full, however many digits it has. Nothing gets rounded away.
2. Everything else is rounded to a consistent 6 significant figures. Not 6 decimal places (which would show wildly different precision depending on a number's size), but 6 meaningful digits, wherever they fall. So 3.280839895… (meters to feet) becomes 3.28084, and 7.840486111… (a fuel-economy conversion) becomes 7.84049. This includes conversion factors that are exact but have more than 6 digits of their own — they get the same consistent treatment as everything else, rather than some numbers getting special-cased to show more digits than others for reasons that would be hard to explain simply.
| What the math produces | What's shown | Why |
|---|---|---|
| 3.280839895013… | 3.28084 | Rounded to 6 significant figures |
| 24 (exactly) | 24 | Exact whole number — shown in full |
| 1,125,899,906,842,624 (exactly) | 1,125,899,906,842,624 | Exact whole number, however many digits |
| 0.1 + 0.2 (float noise) | 0.3 | Recognized as "really" whole once noise is cleaned up |
A genuine floating-point quirk, and why it isn't a bug
Every calculator, spreadsheet, and programming language that uses standard binary floating-point arithmetic (which is nearly all of them) shares a well-documented quirk: a decimal like 1.005 can't be stored exactly in binary, so it's actually held as a value very slightly less than 1.005. Round that to 2 decimal places by hand and you'd expect 1.01; ask a computer, and you'll often get 1.00 instead. It's inconsistent, too — 1.015 rounds correctly the "expected" way, 1.005 doesn't.
This is a property of how binary computers represent decimal fractions, not a flaw specific to this site or any rounding choice we've made — and it only ever affects the very last digit of a number landing almost exactly on a rounding boundary. It's included here because we'd rather tell you about it than have you wonder why a result looks slightly off in a rare edge case.
Thousands separators, and why they don't change with your device's settings
Numbers here use a comma every three digits (1,234,567) and a period for the decimal point — the standard convention for English-language countries including the US, UK, Canada, and Australia. Both the amount you type and the result shown update live with these separators, so a mistyped extra zero is easy to spot at a glance.
Some regions do it differently: much of continental Europe uses a period or space for thousands and a comma for the decimal point (so "1.234,56" or "1 234,56" means what "1,234.56" means here) — and your browser or operating system may even be configured for one of these conventions. We deliberately keep number formatting fixed to the English convention across this English-language site, rather than silently adapting to your device's regional settings, since mixing English text with a different region's number punctuation would be more confusing than consistent, not less. When this site adds other languages, each language's pages will use that language's own numeric convention.
Two ways to see very large or very small numbers
Once a result gets into the millions (or down into tiny fractions), reading a long string of digits gets harder, not easier — so results in that range show a second, smaller line offering the same number in scientific notation, like 1.1259 × 10¹⁵ for a huge byte count. It's written with an actual multiplication sign and superscript exponent rather than calculator-style "1.1259e+15", specifically so it reads naturally even if you don't work with scientific notation often.
For truly extreme results (past roughly a quintillion, or below a billionth), the roles flip: scientific notation becomes the main display, since a fully-expanded decimal would be dozens of digits long, and the plain expanded number is offered as the secondary line instead.
Copying a result
The Copy button next to every result copies the plain number — without thousands separators — along with its unit. Pasting a number with commas into a spreadsheet formula or another program can cause it to be misread or rejected outright, so what gets copied is always the clean, calculation-ready version, even though the display on screen includes separators for readability.