What “accurate” means on this website
A result is considered accurate only within the documented input syntax, number domain, and problem family. The solver should identify the correct class, apply valid transformations, return the complete supported answer, and provide appropriate verification evidence. A visually polished answer without those conditions is not enough.
No claim of universal or perfect mathematical accuracy is made. Software defects, ambiguous input, browser problems, hosting failures, and untested edge cases can occur. Important academic or professional work should be checked independently.
Supported problem families
| Problem family | Current support |
|---|---|
| Linear equations in x | Integer, decimal, and fractional coefficients; variable terms may appear on both sides; one, none, or infinitely many solutions. |
| Quadratic equations in x | Standard-form classification, discriminant analysis, rational factoring where available, exact radicals, and real or complex roots. |
| Linear inequalities in x | First-degree inequalities with explicit reversal when multiplying or dividing by a negative value. |
| Linear absolute-value equations | An isolated supported absolute-value expression with positive, zero, or negative target analysis and complete case checking. |
| Polynomial expressions | Simplification and expansion through the supported parser range, combining matching powers, and rational factorization for recognized cases. |
| Exact fraction arithmetic | Numeric expressions and direct addition, subtraction, multiplication, or division of fractions with reduction and visual models. |
| Least common multiples and greatest common divisors | Two to eight positive integers with prime factorizations, maximum and minimum shared exponents, and exact divisibility checks. |
| Ratio simplification | Two to six nonnegative whole-number, fraction, or decimal parts reduced to an exact whole-number ratio. |
| Polynomial equations and zeros | Degree 3 to 6 through the rational root theorem, synthetic division, the quadratic formula, and exact nth roots; other roots shown as labelled decimal approximations. |
| Inequalities beyond linear | Compound linear inequalities and quadratic or polynomial inequalities with a sign chart and interval notation. |
| Systems of linear equations | Two to four equations with up to four unknowns by elimination, including no-solution and infinitely-many cases. |
| Polynomial division | Long division by any polynomial divisor and synthetic division by a linear divisor, with a rebuild check. |
| Quadratic forms and lines | Completing the square, vertex form, the discriminant, slope from two points or an equation, and line forms. |
| Number conversions | Decimal to fraction (including repeating decimals), fraction to decimal, radical simplification, GCF, and prime factorization. |
| Polynomial evaluation and graphing | Exact substitution for the variable, and graphs up to degree 6 with intercepts and vertex labels. |
How to interpret a verification label
“Verified” means the engine produced candidate-level or equivalence evidence and used the relevant problem classification to establish completeness within the supported domain. It does not mean a human teacher reviewed that individual submission.
“Equivalent” is used for simplification, factoring, or expansion when exact polynomial comparison confirms matching coefficients. “Exact” identifies rational arithmetic or substitution completed without floating-point rounding. “Function checked” means the graph is rendered from exact normalized coefficients, while the visible curve itself remains a sampled display. “Review needed” means complete independent evidence was unavailable.
Problems outside the current release
The site does not solve trigonometric equations, calculus, rational inequalities, equations with more than one square root, nonlinear systems, or polynomials above degree 6. Exponential equations whose answer is not an exact power are given as a logarithm with a 12-digit decimal. Roots of cubic and higher equations that have no exact rational or simple radical form are shown as decimals rounded to six places and labelled as approximate.
An unsupported message is a capability boundary, not a statement that the problem has no solution. Use a method-specific educational resource or qualified instructor for those problem types.
Input interpretation and user checks
Parentheses, fraction bars, exponent notation, and relation symbols affect meaning. Review the “Interpreted as” line before relying on a result. For example, 1/2x may be read differently from 1/(2x), so grouping is essential when the intended structure could be ambiguous.
After solving, compare the normalized problem with the original source and inspect the first step. Substitution is useful only when it checks the problem you meant to enter.
Reporting a suspected error
A useful report includes the exact input, selected operation, number domain, displayed answer, the first step believed to be wrong, expected result, device, and browser. Do not include a student name, school record, login credential, or other personal information.
Confirmed correctness defects should receive priority over cosmetic issues, and material corrections should be described in the public changelog when appropriate. The current contact limitation is explained on the Contact page.