How this calculator reads what you type

Kalkuli works on the whole expression rather than one number at a time, so you can type a calculation exactly as it's written on paper and see a live preview of the answer above the display before you press =. It follows the normal order of operations: brackets, then powers and roots, then multiplication and division, then addition and subtraction.

Functions open a bracket for you. Pressing sin types sin(. You don't have to close the last bracket; Kalkuli closes any that are left open when you press =.
You can skip the × sign in front of a bracket, a function or a constant: 2π, 3(4 + 1) and 2sin(30) all work.
Powers bind tighter than a minus sign, as in standard maths notation: −2² is −4. Write (−2)² if you mean 4.
% means "divide by 100", so 200 × 15% is 30. It doesn't do the shop-till trick where 200 + 15% means 230.
Very large and small numbers use E notation: 6.02214076E23 means 6.02214076 × 10²³. Use the book icon at the top of the page to insert physical constants.
After =, carry on. Pressing an operator continues from the answer; typing a number starts a new calculation. Click any line in History to reuse its result.

Keyboard: digits, + − * / ^ ( ) ! %, Enter for equals, Backspace to delete and Esc to clear. You can type function names too: sin(30), sqrt(2), log(100), pi.

The most common mistake

If sin, cos or tan give a strange answer, check the DEG/RAD button. In radian mode, sin(30) is −0.988, not 0.5, because 30 radians is about 1,719°. I wrote a whole guide on this: degrees vs radians, and how to tell which one you need.

Worked examples you can type in

Each example shows exactly what to type. Copy it into the calculator to check the result.

1. How high does a ball go?

A ball kicked at 25 m/s at 45° to the ground reaches a maximum height of h = v² sin²θ / 2g, ignoring air resistance. With standard gravity g = 9.80665 m/s²:

25² × sin(45)² ÷ (2 × 9.80665) = 15.9331 m

Because sin 45° = √2/2, sin² 45° is exactly ½, so this is just 625 ÷ 39.2266.

2. How many decibels louder?

An amplifier putting out 50 W compared with a 1 mW reference: L = 10 log₁₀(P / P₀).

10 × log(50 ÷ 0.001) = 46.9897 dB

More on why log scales are used like this in logarithms in real life.

3. Continuously compounded interest

£10,000 at 5% a year, compounded continuously for 7 years: A = P e^rt. The eˣ key types e^(.

10000 × e^(0.05 × 7) = 14190.6754859

That's £14,190.68, only 34p more than compounding daily would give, which is why continuous compounding is mostly used as a mathematical shortcut rather than by actual banks.

4. Choosing 3 from 8

The number of ways to pick 3 items from 8 when order doesn't matter is 8! / (3! × 5!).

8! ÷ (3! × 5!) = 56

For bigger numbers use the nCr tool on the statistics calculator, which works with exact whole numbers rather than floating point.

How accurate are the answers?

The arithmetic runs in your browser using 64-bit floating-point numbers (IEEE-754 double precision), which hold about 15–16 significant digits. Kalkuli shows results to 12 significant figures. That hides the binary rounding noise that makes some software print 0.1 + 0.2 as 0.30000000000000004, while keeping far more precision than any real measurement needs.

In degree mode, sin, cos and tan give exact results at whole multiples of 30° and 45° where the true answer is 0, ±½ or ±1, so sin(180) is 0 rather than 1.2 × 10⁻¹⁶, and tan(90) tells you it's undefined instead of returning a huge number. The precision guide explains where floating-point error comes from and when it can bite.