VerifySolveSolve a problem

Matrix Calculator — every operation, proven

Type an operation — [[1, 2], [3, 4]] * [[5, 6], [7, 8]] or det([[6, 1, 1], [4, -2, 5], [2, 8, 7]]) — and this matrix calculator returns the answer with full steps and a verification proof: products rechecked through the transpose identity, determinants computed by two routes, inverses proved by multiplying back to the identity.

  • ✓100% free — no paywall, ever
  • ✓No sign-up, no app
  • ✓Every answer verified before you see it

In short

Matrices hold numbers in rows and columns, and the operations on them follow rules that punish guessing: multiply in the wrong order and the answer changes, invert a singular matrix and there is no answer at all. This matrix calculator computes the core operations on matrices up to 4×4 — add, subtract, scalar-multiply, transpose, trace, multiply, determinant and inverse — with exact fractions throughout and every answer recomputed a second way and proved.

Four ideas carry this page

Addition and multiplication, the determinant, and the inverse. The solver applies all four in the steps — and this matrix calculator refuses honestly when the shapes do not fit or no inverse exists, instead of guessing.

Addition and scalar multiplication

Cij=Aij+Bij(kA)ij=k⋅AijC_{ij} = A_{ij} + B_{ij} \qquad (kA)_{ij} = k \cdot A_{ij}

Multiplication — rows by columns

Cij=∑kAikBkj(m×n)(n×p)=m×pC_{ij} = \sum_{k} A_{ik} B_{kj} \qquad (m \times n)(n \times p) = m \times p

Determinant — 2×2

det⁡(A)=ad−bcfor A=(abcd)\det(A) = ad - bc \qquad \text{for } A = \begin{pmatrix} a & b \\ c & d \end{pmatrix}

Inverse — adjugate over determinant

A−1=1det⁡(A)adj⁡(A)A×A−1=IA^{-1} = \frac{1}{\det(A)} \operatorname{adj}(A) \qquad A \times A^{-1} = I

Worked examples — a product, an inverse, a determinant, a singular matrix

Two matrices multiplied with every dot product shown, an inverse with fractions proved by the identity, a 3×3 determinant computed two ways, and a matrix with honestly no inverse. Each check below is the same recomputation this matrix calculator prints on every problem.

[[1, 2], [3, 4]] * [[5, 6], [7, 8]]

  1. 1.Shapes: 2×2 times 2×2 — inner dimensions match, product is 2×2
  2. 2.C[1,1] = 1×5 + 2×7 = 19; C[1,2] = 1×6 + 2×8 = 22
  3. 3.C[2,1] = 3×5 + 4×7 = 43; C[2,2] = 3×6 + 4×8 = 50

[[19, 22], [43, 50]]

Proof: transpose(C) recomputed as transpose(B) × transpose(A) ✓ — and the reversed product [[23, 34], [31, 46]] shows order matters.

inverse([[1, 2], [3, 4]])

  1. 1.Determinant first: det = 1×4 − 2×3 = −2 — non-zero, so the inverse exists
  2. 2.Cofactor matrix: [[4, -3], [-2, 1]]; adjugate (transposed): [[4, -2], [-3, 1]]
  3. 3.Divide by det: (1/−2) × [[4, -2], [-3, 1]] = [[-2, 1], [3/2, -1/2]]

[[-2, 1], [3/2, -1/2]]

Proof: A × inverse(A) = [[1, 0], [0, 1]] ✓ — the check this matrix calculator prints on every inverse.

det([[6, 1, 1], [4, -2, 5], [2, 8, 7]])

  1. 1.Bareiss elimination, stage 1 (pivot 6, divisor 1): [[6, 1, 1], [0, -16, 26], [0, 46, 40]]
  2. 2.Stage 2 (pivot −16, divisor 6): the last pivot is −306

det = −306

Proof: cofactor expansion along the first row also gives −306 ✓ — two independent routes, one answer.

inverse([[1, 2], [2, 4]])

  1. 1.Determinant first: det = 1×4 − 2×2 = 0
  2. 2.A matrix with determinant 0 has no inverse — the rows are multiples of each other

no inverse exists

Proof: both determinant routes agree it is 0 — "no inverse" is the answer, not an error, and this matrix calculator says so plainly.

Common mistakes — and how to dodge them

Multiplying in the wrong order

Matrix multiplication is not commutative: AB and BA are different problems with different answers — [[1, 2], [3, 4]] × [[5, 6], [7, 8]] is [[19, 22], [43, 50]], reversed it is [[23, 34], [31, 46]]. Swapping the order silently is the most common matrix error, and it is invisible unless something shows both. This matrix calculator states the shape rule in the steps and shows the reversed product for square pairs, so an order error shows up instead of hiding.

"Dividing" matrices

There is no matrix division. Writing A ÷ B has no meaning — the operation that plays the role of division is multiplying by the inverse: A × inverse(B), and only when B has a non-zero determinant. Asking for a quotient of two matrices here refuses with that explanation, because a page that invented an answer would teach a false operation. If B is singular, there is no division by B at all, and the page says so with the zero determinant shown.

Assuming every square matrix has an inverse

Square is necessary but not enough: a matrix has an inverse exactly when its determinant is non-zero. [[1, 2], [2, 4]] is square, but its second row is twice its first, its determinant is 0, and no matrix multiplies with it to give the identity. Treating "no inverse" as an error to work around — or inventing entries — hides a real mathematical fact. Ask this matrix calculator for the inverse of a singular matrix and it answers "no inverse exists" with the zero determinant shown — the fact, not an error.

Matrix Calculator — questions

What does a matrix calculator do?

It computes with grids of numbers. This one handles the core operations on matrices up to 4×4: addition and subtraction, scalar multiplication, transpose, trace, multiplication, determinant and inverse. Every entry is an exact fraction — no floating-point drift — and every answer is verified by independent recomputation: products rechecked through the transpose identity, determinants computed by two routes, inverses proved by multiplying back to the identity. Vectors, eigenvalues, rank, row reduction and matrix powers are out of scope and refused with the reason named.

How do I type a matrix?

In double brackets, one bracketed row at a time: [[1, 2], [3, 4]] is the 2×2 matrix with rows 1, 2 and 3, 4. Then say which operation: [[1, 2], [3, 4]] * [[5, 6], [7, 8]] multiplies, det([[1, 2], [3, 4]]) takes the determinant, inverse([[1, 2], [3, 4]]) inverts, transpose(...) and trace(...) do what they say. Entries can be integers, decimals or fractions like 1/2. This matrix calculator accepts up to 4 rows and 4 columns, and a leading verb like "calculate" is fine.

Why does my matrix have no inverse?

Because its determinant is 0. A matrix has an inverse exactly when its determinant is non-zero: the inverse is the adjugate divided by the determinant, and division by zero is impossible. Rows that are multiples of each other — [[1, 2], [2, 4]] — always give determinant 0. That is a mathematical fact, not an input error, so the answer here is "no inverse exists" with the zero determinant shown and verified by both determinant routes, not a refusal and not a faked matrix.

Does the order matter in matrix multiplication?

Yes — AB is generally different from BA, and sometimes only one order is even defined. [[1, 2], [3, 4]] × [[5, 6], [7, 8]] is [[19, 22], [43, 50]], but reversed it is [[23, 34], [31, 46]]. The shape rule decides what is allowed: an m×n matrix times an n×p matrix gives m×p, so the inner dimensions must match. The steps state the shapes before multiplying, and for square pairs they show the reversed product too, so an order error shows up instead of hiding.

Does this matrix calculator do vectors, eigenvalues, or systems of equations?

No — and it says so rather than faking it. This page computes the core matrix operations only: it never does vector operations, never computes eigenvalues or rank, never row-reduces, and never raises a matrix to a power. Systems of equations belong to the System of Equations Calculator, which solves them by elimination and substitution. The boundary keeps every answer on this page honest: what it computes, it proves.

Is the matrix calculator free?

Yes — free, no sign-up, no app, full steps and verification included on every problem. The scope is matrices up to 4×4: add, subtract, scalar-multiply, transpose, trace, multiply, determinant and inverse, with exact fractions throughout. Anything outside that scope — vectors, eigenvalues, rank, row reduction, matrix powers — is refused with the reason named instead of answered approximately.

Keep solving

A note on honest use: VerifySolve is built for learning. Use the steps to understand the method, then close the tab and try one on your own — that is where the grade actually comes from. If your teacher or school has a policy on calculators, follow it.