Choose the target and approach direction, then examine the behaviour of an expression near that point.
Limit Solver
Maths solverUse a photo, screenshot, or scanned question.
Supported images · up to 20 MB
JPG, PNG, WebP, GIF, AVIF, BMP, ICO, SVG, HEIC/HEIF, and TIFF. Animated and multipage files use the first image. If a file cannot be opened, export it as JPG or PNG.
Write maths as it looks on paper. Use fractions, roots, powers, and the keys below. Select the main expression or a smaller maths field before using the keys. Optional boxes may be left blank; the solver will ask if essential information is missing. Tap a placeholder or use the box arrows below the keyboard to move between index, limits, and expression fields.
Write an equation or describe a word problem. Include any units and what you need to find.
AI-assisted worked solutions. Check the result against your original question. Please leave out personal information.
Step by Step Calculation
Follow each part of the solution below.
Describe the approach, not just the expression
A limit asks what a function approaches as its input gets close to a chosen value. This calculator separates the expression, variable, approach value, and direction so the question is explicit. Enter the expression in the main maths field, select its variable, and use the smaller approach field for the target. The direction menu distinguishes a two-sided limit from an approach only from the left or only from the right.
At a finite point, left means inputs smaller than the target and right means inputs larger than it. These are directions along the input axis, not descriptions of whether the function’s output is positive or negative. For an approach to positive or negative infinity, use the infinity symbol in the approach field; a finite-point left/right choice is then irrelevant.
Try substitution before choosing a harder method
If a function is continuous at the target, direct substitution gives its limit. A polynomial illustrates this cleanly. As \(x\) approaches two, the expression \(x^2+3x-1\) approaches \(4+6-1=9\). There is no need to invent an indeterminate form or apply a derivative rule.
Substitution is a diagnostic step for fractions too. A nonzero denominator often allows immediate evaluation. A result of \(0/0\), however, is not the answer; it signals that the expression’s local behaviour needs further analysis.
A removable gap can still have a limit
Consider \((x^2-9)/(x-3)\) as \(x\) approaches three. The original expression is undefined at three. Factoring the numerator gives \((x-3)(x+3)\). For nearby inputs with \(x\ne3\), the shared factor cancels, leaving \(x+3\).
This calculation does not assign a value to the original function at three. It shows that its nearby values approach six. Changing a function at a single point can alter its value there without changing its limit, because the limiting question concerns surrounding inputs.
Compare both sides before reporting a two-sided answer
For \(|x|/x\), inputs to the right of zero give one, while inputs to the left give negative one. Since the one-sided results differ, the two-sided limit at zero does not exist.
A calculator should not average these answers to produce zero. Likewise, \(1/x\) tends towards opposite signed infinities on the two sides of zero. A claim that it simply “equals infinity” would hide the directional difference. For a function with an endpoint in its real domain, such as \(\sqrt{x}\) at zero, interpret the approach within the available domain and state that qualification.
Look at growth rates for very large inputs
When the target is infinity, the highest powers often reveal the dominant behaviour of a rational expression. Divide every term of \((3x^2+1)/(2x^2-5)\) by \(x^2\). The terms involving \(1/x^2\) approach zero, leaving the ratio of the leading coefficients:
Infinity is a description of unbounded growth, not a number to substitute into ordinary arithmetic. An expression such as infinity divided by infinity does not determine the answer by itself; different functions can produce very different limits.
Keep exact coefficients when lower-order terms cancel
Some limits depend on a term far beyond the first approximation. For a logarithm containing \(\sin x/x\), replacing sine by just \(x\) loses the information needed after cancellation. The expansion in radians begins:
Adding \(x^2/6\) and \(x^4/180\) cancels the first two terms. Dividing by \(x^6\) leaves a constant of \(-1/2835\), plus terms tending to zero. The “Series cancellation” example lets you inspect this calculation. Every submitted limit is explained by the AI solver through the numbered queue. For supported finite limits at zero, an exact rational-series calculation checks the final result and supplies the AI with a checked expansion. Other expressions rely on the AI explanation and should be checked independently.
Use special rules with their conditions
The standard trigonometric limit \(\sin x/x\to1\) at zero uses radians. L’Hôpital’s rule is also conditional: it applies to suitable differentiable quotients in qualifying indeterminate forms, not to every fraction. Oscillation needs separate attention; \(\sin(1/x)\) keeps moving between values near zero and has no limit there.
Try \((x^2-4)/(x-2)\) at two, \(1/x^2\) at zero, and \(1/x\) at positive infinity. The results are four, positive unbounded growth, and zero, respectively. Compare the symbolic explanation with a graph, but remember that a few sampled points cannot prove a limit. The Algebra Calculator can help with factorisation before you return to the limiting argument.