13 case formats in one text box. Paste, click, copy.
Turn plain words into tiny, raised, mirrored and symbol text you can paste anywhere.
No-fuss tools for age, sleep, time and money, plus a strong password maker.
Round to the nearest whole number, ten or hundredth, to decimal places or significant figures, in 9 rounding modes. Exact decimal math, with the deciding digit shown.
Private by design. Everything runs in your browser and is never uploaded.
Rounding replaces a number with a shorter, nearby value at the precision you choose. Find the rounding digit, look at the digit just to its right, and round up if that digit is 5 or more, otherwise keep it. So 2.675 to two decimal places is 2.68.
This rounding calculator does that for any place value, any number of decimal places, significant figures or the nearest multiple, and it highlights the digit that decided the result. It works on the exact decimal digits you type, so 1.005 to two decimal places really is 1.01. For measurement work with significant-figure rules for adding and multiplying, use the significant figures calculator.
| Number | Rounded to | Result |
|---|---|---|
| 7.5 | Nearest whole number | 8 |
| 3.14159 | 2 decimal places | 3.14 |
| 2.675 | Nearest hundredth | 2.68 |
| 1,847 | Nearest hundred | 1,800 |
| 0.004567 | 2 significant figures | 0.0046 |
| 187 | Nearest 5 | 185 |
| 2.5 | Whole number, half to even | 2 |
| −3.5 | Whole number, half away from zero | −4 |
Four steps. The answer updates as you type, with no Calculate button.
Commas, a leading plus or minus and scientific notation such as 6.02e23 all work. On a phone, the ± button adds or removes the minus sign.
A place value from millions to millionths, a number of decimal places, a number of significant figures, or the nearest multiple such as 5, 0.05 or 1/8.
The standard mode rounds halves away from zero. Switch to half to even for bankers rounding, or to ceiling, floor or truncate when you always need to go one way.
The rounding digit is marked in green and the deciding digit in violet, with the two neighbours and the reason for the choice. Copy the result or a share link.
Round-to-nearest modes only disagree on exact ties such as 2.5. Directed modes (ceiling, floor, truncate) ignore how big the remainder is. Each value below is rounded to a whole number.
| Mode | Rule | 2.5 | 3.5 | −2.5 | 2.4 | −2.6 | −3.5 | Where you meet it |
|---|---|---|---|---|---|---|---|---|
| Half away from zero | Ties move away from zero | 3 | 4 | −3 | 2 | −3 | −4 | School maths, Excel ROUND |
| Half up | Ties move toward +∞ | 3 | 4 | −2 | 2 | −3 | −3 | Many textbooks for positive numbers |
| Half down | Ties move toward −∞ | 2 | 3 | −3 | 2 | −3 | −4 | Rare, mostly for symmetry |
| Half toward zero | Ties move toward zero | 2 | 3 | −2 | 2 | −3 | −3 | Rare, Python ROUND_HALF_DOWN |
| Half to even (bankers) | Ties go to the even neighbour | 2 | 4 | −2 | 2 | −3 | −4 | IEEE 754, Python, statistics, finance |
| Half to odd | Ties go to the odd neighbour | 3 | 3 | −3 | 2 | −3 | −3 | Rare, avoids carrying into a new digit |
| Ceiling | Always up, toward +∞ | 3 | 4 | −2 | 3 | −2 | −3 | Boxes, buses or tiles needed |
| Floor | Always down, toward −∞ | 2 | 3 | −3 | 2 | −3 | −4 | Whole items you can afford |
| Truncate | Drop extra digits, toward zero | 2 | 3 | −2 | 2 | −2 | −3 | Excel ROUNDDOWN, cutting off decimals |
Watch the names: what this page calls half away from zero is called ROUND_HALF_UP in the Python decimal module, while many maths references use half up for ties that go toward positive infinity, so −2.5 becomes −2. For positive numbers the two are the same.
Every place-value rounding follows the same rule. The rounding digit is the one in the place you want to keep; the deciding digit is the next one to the right. Keep or raise the rounding digit, then replace everything to its right with zeros (left of the decimal point) or drop it (right of the point).
The table rounds one number, 4,718.2659, to each place with the standard mode. Rounding to the nearest hundredth is the same as rounding to 2 decimal places, which is how prices and most money amounts are shown; the expense calculator keeps its totals at that precision.
| Round to | Deciding digit | Result |
|---|---|---|
| Thousands | 7 | 5000 |
| Hundreds | 1 | 4700 |
| Tens | 8 | 4720 |
| Whole number (ones) | 2 | 4718 |
| Tenths | 6 | 4718.3 |
| Hundredths | 5 | 4718.27 |
| Thousandths | 9 | 4718.266 |
Checked against the rounding tools that rank for this search in November 2026.
| Feature | This calculator | Typical rounding calculator |
|---|---|---|
| Arithmetic | Exact decimal digits, up to 400 digits | Often binary floating point, so 1.005 can come out as 1.00 |
| Rounding modes | 9, including half to even, half to odd and truncate | From 1 (round half up only) to 9 |
| Precision | Place value, decimal places, significant figures, any multiple or fraction | Usually one or two of these; some accept whole numbers only |
| Working shown | Rounding digit, deciding digit, both neighbours and the reason | Result only |
| All modes at once | Yes, a table for the current number | Rarely |
| Negative numbers | Every mode defined and shown for negatives | Often unclear, and one well-known page shows a sign error |
| Float warning | Shows when ordinary binary rounding would print a different answer | Not offered |
| Lists and sharing | Round a whole list, copy, download, share link with your settings | Usually one number at a time |
Round to the nearest ten, hundred or tenth and see exactly which digit made the difference, so you can check each step of your own working.
Round to 2 decimal places for cents, or to the nearest 0.05 where small coins are no longer used. Half to even avoids a steady upward drift across many amounts.
Round a measurement to the number of significant figures your instrument supports, with scientific notation shown when trailing zeros would be unclear.
Round 4,718 to 4,700 or 5,000 to estimate a total quickly, then compare the estimate with the exact figure.
Ceiling tells you how many boxes, buses or packs you need, and floor tells you how many whole items fit a budget. For hours and minutes, the time calculator adds and subtracts durations.
Check what Excel ROUND, Python decimal or a JavaScript toFixed call should return, and paste a whole column into the list box to round it in one go.
Most rounding surprises come from one of these six points. To keep a set of rounded results, paste them into the online notepad, or browse the other free online calculators on this site.
2.675 is stored as 2.67499999999999982236431605997495353221893310546875, so plain float rounding gives 2.67. This calculator rounds the digits you typed instead.
Rounding in stages can be wrong: 2.449 goes to 2.4 in one step, but to 2.45 and then 2.5 in two.
Every round-to-nearest mode gives 2.68 for 2.6751. They differ only when the rest is exactly half, such as 2.675 to two places.
Half up and half away from zero give different answers for −2.5 (−2 and −3), and floor differs from truncate for −2.7 (−3 and −2).
2.50 to 2 decimal places says the value is known to the hundredth. The calculator keeps those zeros, and drops a minus sign from a zero result.
Leading zeros never count, so 0.0004567 to 2 significant figures is 0.00046. Trailing zeros in 1,200 are ambiguous; write 1.2 × 103 to show two.
Numbers can have up to 400 digits and an exponent between −1000 and 1000, and precision goes up to 100 decimal places or significant figures. The list box reads the first 200,000 characters.
It rounds one value at a time. It does not apply the significant-figure rules for sums and products of measurements; the significant figures calculator linked above covers those.
Tax, invoicing and lab rules can require a specific mode or precision. Check the rule that applies to you, such as your tax authority or the Excel ROUND documentation if you need to match a spreadsheet.
The mode definitions and examples on this page were checked against these references on November 9, 2026. The float values were computed from the IEEE 754 double that each decimal is stored as.
Short answers to what people ask most about rounding.
Look at the tenths digit, the first digit after the decimal point. If it is 5 or more, add one to the ones digit; if it is 4 or less, keep the ones digit, then drop everything after the point. So 7.5 becomes 8 and 7.49 becomes 7.
Look at the third digit after the decimal point. If it is 5 or more, the second digit goes up by one; otherwise it stays, so 3.14159 becomes 3.14 and 2.675 becomes 2.68.
They differ only for negative numbers that sit exactly halfway. Half up sends -2.5 to -2, toward positive infinity, while half away from zero sends it to -3. For positive numbers both give the same answer.
Bankers rounding, also called round half to even, sends an exact tie to the neighbour whose last digit is even, so 2.5 becomes 2 and 3.5 becomes 4. Ties go up half the time and down half the time, which removes the upward drift of always rounding 5 up. It is the default rounding in IEEE 754 floating point and in the Python decimal module.
Binary floating point cannot store 2.675 exactly; the nearest value it can store is 2.67499999999999982236431605997495353221893310546875, which is just below the halfway point. Programs that round that stored value print 2.67. This calculator rounds the decimal digits you typed, so it gives 2.68.
In standard rounding you round the size of the number and keep the minus sign, so -3.7 becomes -4 and -3.2 becomes -3. Modes that follow a direction behave differently: the floor of -2.3 is -3 and the ceiling of -2.3 is -2.
Truncating drops the extra digits, which always moves toward zero, while floor (round down) always moves toward negative infinity. They agree for positive numbers but not for negative ones: truncating -2.7 gives -2 and flooring it gives -3.
Choose Nearest multiple and type the step, or tap one of the quick buttons. The calculator divides by the step, rounds the quotient to a whole number and multiplies back, so 187 to the nearest 5 is 185 and 2.675 to the nearest 0.05 is 2.70.
Decimal places count digits after the decimal point, while significant figures count digits from the first non-zero digit. For example, 0.004567 to 2 decimal places is 0.00, but to 2 significant figures it is 0.0046.
No. Round the original number once, straight to the precision you need. Rounding 2.449 to one decimal place gives 2.4, but rounding it first to 2.45 and then to one decimal place gives 2.5, which is wrong.
No. The rounding runs in your browser and nothing you type is sent to a server. The share link keeps your number in the part of the address after the # sign, which browsers do not send to the website.