SciCalc.net
IndiaFree tools

Definite integral calculator

Type a function of x and two limits to get the definite integral, the signed area under the curve, to 10 significant figures with an error estimate. It works the number out numerically; it does not give the antiderivative formula.

Last updated

∫ from 0 to 3 of x2 dx =

9

Estimated error: ± 1.3E−15

Working

  1. 1Limits: from 0 to 3 = 0 to 3
  2. 2Adaptive 15-point Gauss–Kronrod rule: the interval is split where the function changes fastest = 1 piece, 15 function valueseach piece compares a 7-point and a 15-point rule to estimate its error
  3. 3Add the pieces = 9estimated error ± 1.3E−15
Try:

How it works

The definite integral ∫ from a to b of f(x) dx is the signed area between the curve and the x-axis: area above the axis counts as positive, area below as negative. When an antiderivative F is known, the fundamental theorem of calculus gives it exactly:

∫ from a to b of f(x) dx = F(b) − F(a)

This calculator finds the number directly. It uses an adaptive Gauss–Kronrod rule: on each piece of the interval it compares a 7-point and a 15-point weighted sum, and keeps splitting the pieces with the largest disagreement until the total error estimate is about 10−12 of the answer. That works for functions with no simple antiderivative, such as e−x².

Worked example

∫ from 0 to 3 of x² dx: an antiderivative is x³/3, so the answer is 3³/3 − 0³/3 = 27/3 = 9.

∫ from −1 to 1 of √(1 − x²) dx is the area of a half circle of radius 1: π/2 = 1.570796327.

Tips

  • Swap the limits and the answer changes sign.
  • If the function has no value somewhere between the limits (such as 1/x at 0), the calculator says so instead of giving a number: that integral is improper and needs a limit argument.
  • For an antiderivative formula, use the power rule backwards: ∫ xⁿ dx = xⁿ⁺¹/(n + 1) + C, for n ≠ −1.

Questions

Does this integral calculator give the antiderivative?

No. It gives the value of a definite integral (with two limits) as a decimal, to 10 significant figures. It does not find the formula for an indefinite integral.

How accurate is it?

It stops when its own error estimate is about one part in 10¹², and shows 10 significant figures. It is checked against a 50-digit reference in our test suite. Functions that blow up at a limit, such as 1/√x at 0, are handled but take more steps.

What does a negative answer mean?

More of the curve lies below the x-axis than above it between the limits, or the upper limit is smaller than the lower one.

Can it integrate to infinity?

No. Both limits must be finite numbers. For an integral to infinity, try larger and larger upper limits and watch whether the answer settles.

Sources

  1. NIST Digital Library of Mathematical Functions, §1.4 Calculus of One Variable. §1.4(v) Definite Integrals: Riemann-sum definition 1.4.18; swapping the limits changes the sign 1.4.20; improper integrals 1.4.22–1.4.23; Fundamental Theorem of Calculus 1.4.26, ∫ from a to b of f = F(b) − F(a). Power rule ∫xⁿ dx = xⁿ⁺¹/(n + 1) + C, n ≠ −1: §1.4(iv), 1.4.17. Checked 22 November 2026.
  2. QUADPACK routine DQK15 (Netlib). Prologue and data statements: 15-point Kronrod rule formed by optimal addition of abscissae to the 7-point Gauss rule; nodes xgk and weights wgk, wg; error estimate starts from |resk − resg| × half-length. Checked 22 November 2026.
  3. QUADPACK readme (Netlib). Description of QAG: globally adaptive integrator using pairs of Gauss–Kronrod quadrature formulae. Checked 22 November 2026.
  4. NIST Digital Library of Mathematical Functions, §3.5 Quadrature. §3.5(v) Gauss Quadrature: nodes are zeros of orthogonal polynomials, weights 3.5.18; Gauss–Legendre formula on [−1, 1] with w(x) = 1, 3.5.21 (the 7-point Gauss rule inside the 7–15 pair). Checked 22 November 2026.