How to use Matrix Calculator
- Choose the operation. Add, Subtract and Multiply use both A and B; Transpose, Determinant, Inverse and Rank use A only.
- Enter one matrix row per line using commas or spaces between numbers, or edit the synchronized grid. Keep every row the same length.
- For addition/subtraction, match both matrix dimensions. For multiplication, A’s column count must equal B’s row count. Determinant and inverse require square A.
- Calculate and inspect the result dimensions and numerical threshold. Inverse and determinant working is listed only for matrices up to 4 by 4.
Example: Matrix Calculator
Invert the 2 by 2 matrix with rows [1, 2] and [3, 4]. Matrix B is not used for this operation.
Options
- Matrix shape
- Each matrix may have 1 to 6 rows and columns. Addition and subtraction need the same shape. If A is r by k and B is k by c, multiplication produces r by c. Operand order matters.
- Pivot threshold
- Rank, inverse and related operations use the stated numerical threshold. Small pivots can be treated as zero.
Supported inputs and limits
- Each matrix has 1 to 6 rows and columns, with at most 4,000 input characters.
- Enter plain decimal values of magnitude at most 10¹², with up to 60 characters, 40 digits and 20 fractional places per cell; exponent notation is refused.
- The grid and pasted rows stay synchronized.
- Working uses floating-point arithmetic.
- Pivots at or below 10⁻¹⁰ of the largest original entry count as zero, so nearly dependent matrices can change rank or fail inversion after rounding.
More details
Determinant and inverse need square matrices.
Step lists are provided up to 4 by 4; larger accepted matrices return numerical results without the full working.
There is no symbolic algebra or exact rational guarantee.
Where your input is processed
This tool does not upload your inputs. See the limits for anything saved on your device.
A refusal can reflect numerical sensitivity
Nearly dependent rows can make an inverse unstable even when an exact symbolic inverse exists. Small changes in scale or entered rounding can alter a numerical rank. Check the input precision and threshold before interpreting the result as a statement about an exact mathematical matrix.
Questions about Matrix Calculator
Why must matrix dimensions match for addition?
Each entry is added to the entry at the same position, so both matrices need the same row and column counts.
Can every square matrix be inverted?
No. A singular matrix has no inverse. Values near numerical singularity can also make a floating-point inverse unreliable.
Is a computed rank exact?
The reported numeric rank uses a pivot threshold. Very small perturbations can affect rank for nearly dependent rows or columns.