Scientific Calculator

Stay in Scientific until the input stops being one expression

Scientific is the default tab and handles anything you can write on one line — trigonometry, logarithms, powers, roots, factorials and percentages, well beyond the four basic operations. Switch tabs only when the shape of the input changes: a list of numbers belongs in Statistics, a pair of fractions in Fractions, a named physical quantity in Constants. Each panel hands its answer back with a Use button, so the main keypad stays where every final number is assembled.

1. Enter an expression. Tap the keypad or click the calculator and type with your keyboard — digits 0–9, + − * / ^, parentheses, % and !. The result previews live as you type; press = or Enter to lock it in.
2. Switch to the Scientific keypad for functions. Use the Scientific tab for sin, cos, tan, ln, log, powers (xʸ and x²), roots (√), reciprocals (1/x) and factorials (n!). Tap 2nd to reach inverse functions such as sin⁻¹, eˣ and 10ˣ.
3. Choose degrees or radians. Trigonometry uses degrees by default. Switch to RAD for radian input; the unit badge above the result shows which is active so you never mix them up by accident.
4. Use memory and history. M+ adds the shown value to memory and MR recalls it; MC clears it. Every result is saved to History on this device — tap an entry to reuse it, copy individual values, search history, or export as CSV.
5. Explore Statistics, Fractions, and Constants. Switch to Statistics to compute descriptive stats from a data list. Switch to Fractions for fraction arithmetic and decimal-to-fraction conversion. Switch to Constants to browse and insert 15 physical constants.

There is no sixth tab for algebra: this is a numeric calculator, not a computer algebra system (CAS). It does not solve equations, simplify algebra, differentiate, integrate, handle matrices, or show step-by-step symbolic work — for those, use a CAS such as Wolfram Alpha or SymPy. Nothing is downloaded either: it runs entirely in your browser, behaves the same on desktop and mobile, and saves history and memory locally on your device.

Typing reaches the operators; sin, log and π need the keypad

2^3^2 is 512, −3² is −9, and 200 + 10% is 220

Expressions follow standard precedence (PEMDAS/BODMAS): parentheses, exponents, then ×/÷, then +/−. Exponents are right-associative, so 2^3^2 means 2^(3^2) and returns 512 rather than 64; and a leading minus binds looser than a power, so −3² is −(3²) = −9 while 2^−3 is 0.125. Implicit multiplication needs no × sign — 2π, 2(3+1) and 3sin(30) all evaluate.

A quick reference to the functions you will use most. Every one has an on-screen button and evaluates live in the calculator above.

Degrees vs radians (DEG / RAD)

Degrees split a circle into 360 parts; radians use the arc length of a unit circle, so a full circle is 2π. DEG is the default — everyday geometry, where sin(30) = 0.5 — and RAD suits calculus and physics, where the same 0.5 comes from sin(π/6). The badge above every result shows the active unit, and sin⁻¹, cos⁻¹ and tan⁻¹ answer in that same unit, so if sin(30) does not read 0.5 the unit is the first thing to check, not the arithmetic.

How percentages are interpreted

A standalone percentage is just a fraction: x% means x ÷ 100, so 50% = 0.5. In addition and subtraction it turns contextual, taking a percentage of the left-hand value — 200 + 10% = 220, 200 − 10% = 180 — which matches how shops and most desktop calculators apply discounts and markups. With × or ÷ it stays a plain fraction: 5 × 10% = 0.5.

Five expressions that show the engine is behaving

Each of these evaluates live in the calculator above — tap an example chip to load it. They show how the engine handles angles, logarithms, powers, factorials, percentages, implicit multiplication, and scientific notation. This is an independent browser-based calculator, not a replica of the Casio fx-991 series or of Desmos’s graphing-first interface: the button layout and features differ, but the core math does not, so these five rows cross-check the calculator on your desk. If a row disagrees, check the angle unit and the percent convention before the arithmetic.

Population σ or sample s: the n−1 choice is yours, not the panel’s

Switch to the Statistics tab to compute descriptive statistics from a list of values. Paste numbers from a spreadsheet or type them manually, separated by commas, spaces, or new lines.

The panel calculates:

• Count, Sum, Mean, Median — core summary measures.
• Min, Max, Range — the spread of your data.
• Population standard deviation (σ) and variance (σ²) — when your list is the entire population.
• Sample standard deviation (s) and variance (s²) — when your list is a sample (uses n−1 in the denominator; requires at least 2 values).

Both spreads are shown at once and the panel does not pick one for you: read σ when the list is the whole population, and s when it is a sample drawn from a larger one. Dividing by n−1 rather than n makes s the larger of the two, and the gap is widest on short lists. Click Use beside any result to insert it into the main calculator — the way to chain a standard deviation into a further expression.

Scope: This is a numeric descriptive-statistics tool. It does not perform hypothesis tests, confidence intervals, regression, or inference. Results are rounded by the current display precision.

0.142857 comes back as 1/7; 0.333 stops at 333/1000

The Fractions tab has two tools:

• Fraction arithmetic — enter any two fractions (numerator and denominator separately) and choose + − × ÷. The result is shown fully reduced and as a decimal. Mixed-number notation is used for improper fractions (e.g. 7/4 → 1 3/4).
• Decimal → Fraction — enter a decimal (like 0.333 or 0.142857) to find the nearest fraction with a denominator of 1,000 or less. Useful for recognising recurring decimals and confirming division results.

The conversion is a search: it tries every denominator up to 1,000 and keeps the closest match. So 0.142857 resolves to 1/7, while 0.333 — which the cap hits exactly — stays 333/1000 rather than rounding up to a third. Beyond that cap you get the nearest approximation, not an exact conversion. All fractions are reduced to lowest terms using the greatest common divisor. Click Use decimal in calculator to continue with a result in the main expression.

The constants panel inserts a magnitude and drops the unit

The Constants tab provides quick access to 15 physical and mathematical constants from the NIST CODATA 2018 database. Search by name, symbol, or category (Math, Physics, Chemistry).

Available constants include π, e, the golden ratio φ, speed of light c, gravitational acceleration g, Planck constant h, Avogadro number Nₐ, Boltzmann constant k_B, elementary charge q, gas constant R, Stefan-Boltzmann constant σ, gravitational constant G, proton mass mₚ, electron mass mₑ, and the fine-structure constant α.

Click Use to insert any constant into the current expression, then switch to the Scientific or Basic tab to continue. Only the numeric magnitude travels with it: unit tracking is not performed, so the calculator cannot warn you when metres meet centimetres — a mixed-unit expression returns a confident, wrong number. Keep your units consistent.

tan(90°) reads “undefined” rather than a huge finite number

The tangent of 90° (and 270°) is undefined — its graph has a vertical asymptote there. Because 90° is not exactly representable in binary, a naïve calculation would return a huge but finite value that misrepresents the truth, so the calculator reports it as undefined instead. The same refusal covers √ of a negative number, ln of zero, and dividing by zero: every expression it cannot evaluate gets a short, plain-language message in the result area (announced to screen readers too) rather than a wrong number. The usual cases:

15–17 digits computed, 12 shown, and n! stops at 170

Every result is an IEEE-754 double: about 15–17 significant digits, of which the Auto display shows up to 12 significant figures. That gap is deliberate — it absorbs the tiny rounding artefacts of binary arithmetic, so the identity cos(90°) reads 0 rather than 6×10⁻¹⁷. You can fix the output to 2, 4, 6, or 8 decimals instead. Three limitations follow:

• Trailing digits can drift. A few of the last digits may differ from an exact hand calculation, and very large or very small numbers switch to ×10ⁿ scientific notation.
• Factorials end at 170. Beyond that, n! overflows to infinity — the format cannot hold the number, so the calculator does not pretend it can.
• Double-check anything consequential. Verify figures that inform important academic, financial, or engineering decisions. These limits come from binary floating point, not from this engine.

Related calculators

More in Math, or browse all calculators.

Sources & method

• Wolfram MathWorld — order of operations & functions
• NIST CODATA 2018 — physical constants
• How we calculate — methodology

Method. Every result is computed in your browser by a hand-rolled expression engine (tokeniser → parser → evaluator) with no third-party math library, validated against hand-checked cases on every change. See the precision ceiling for rounding details and the rest of this engine’s limitations.