📖 How to Play Calcudoku: Math Wizard
PAUSED
Start a new game?
Your current progress will be lost.
Fill the grid so every row and every column contains each number exactly once — that part is Sudoku. The twist is the cages: the grid is divided into outlined groups, each marked with a target and an operation, and the numbers inside a cage must combine to hit that target. A cage marked "6×" needs numbers that multiply to six. So you're solving two puzzles at once — the row-and-column logic and the arithmetic — and they constrain each other.
There's a full Story mode with thirty levels and wizard ranks to climb, seven difficulty levels across grid sizes, and a fresh daily challenge each midnight. Send a friend the exact grid with the challenge link.
Where Sudoku gives you starting digits, Calcudoku gives you none — every deduction starts from the cage arithmetic alone. The seven difficulty levels, Beginner up to Cyborg, change both grid size and how forgiving the cage operations are. The daily grid is identical for every player, and a Challenge link reproduces your exact puzzle for whoever you send it to.
I was in love with math as a kid. Not the tidy classroom kind — the kind where a number puzzle grabs you after school and doesn't let go until it's finished. Calcudoku is the puzzle that let me enjoy the thing I was actually good at, and that's the spirit I tried to build into this version: a math puzzle that rewards the person who likes numbers, rather than punishing the person who doesn't.
While testing this build I clocked my personal best on a 6×6 medium at 1:17. I'm telling you that partly out of vanity and partly because it's a useful benchmark — under ninety seconds is very achievable on that size once the cage arithmetic becomes reflex instead of calculation. The path from five minutes down to under two is the whole pleasure of this game, and it's shorter than most players expect.
The trap I see players fall into constantly involves subtraction and division cages. Both are order-independent, and half the people who write to us about being "stuck" are stuck on exactly this. A two-cell cage marked 2− can be 5 and 3 or 3 and 5 — the puzzle doesn't care which cell holds which. A 3÷ cage is 6 and 2 in either order. Players fixate on one arrangement, pencil it in mentally, and stall when it conflicts three moves later. The fix is a habit: hold both operand orders alive in your head until a neighbouring cage or a row conflict forces one of them. The moment you treat a subtraction cage as directional, you've invented a rule the game doesn't have.
Multiplication cages deserve one more note, because they're where number-love pays off. A cage marked 6× in a small grid has very few factorisations — and in a 4×4, where only the digits 1 through 4 exist, a 12× cage over two cells can only be 3 and 4. Factoring the target before touching the grid often solves a cage outright. This is the deduction the game is really about: the row-and-column logic and the arithmetic constraining each other until only one answer survives.
My honest recommendation for a first session: play a 4×4 on an easy setting and don't write anything until you've factored every multiplication cage in your head. Then move up a size. Story mode will carry you through the ranks at a fair curve — I tuned it so each chapter teaches one habit — but the daily is where the game lives for me. Same grid for everyone, every midnight, scores honestly comparable. Beat 1:17 on a 6×6 medium and you've beaten the person who built it, which I promise is more satisfying than it should be.
Don't start with the biggest cage — start with the most constrained one. A single-cell cage just hands you its number outright, so fill those first. Then look for cages where the math only works one way: a two-cell cage targeting a high product, or a small subtraction, often has just one possible pair. Lock those in and the row-and-column rules cascade from there.
The real engine of the puzzle is the back-and-forth between the two rules. A cage tells you a cell is a 2 or a 4; the column tells you the 4 is already taken; so the cage cell must be the 2. I'm constantly bouncing between "what does the arithmetic allow" and "what does the row already forbid," and the answer falls out of the overlap. Pencil in the candidates for the tight cages, then let the two constraints eliminate each other.
Calcudoku is the puzzle I lost ten days to as a kid, and it's still the one I'd point to for teaching real reasoning — because it makes arithmetic and logic work together instead of separately. I found this in an old book when I was seven and disappeared into it for ten days straight — and something clicked. It made me obsessed with numbers in a way that never wore off; I ended up keeping a little book of mathematical proofs I'd worked out myself, including my own shortcut for telling whether a number is divisible by 13. (I'll share that proof on the blog — and if anyone finds a number that breaks it, tell me, because I'd genuinely love to be proven wrong and learn something.) This is the game that started all of that.
Every solved cell is a tiny proof: the cage allows this, the row forbids that, so the answer must be this. That's deduction, the same move that underpins everything from algebra to debugging, and the larger grids stretch it by stacking more constraints on top of each other. No download, no sign-up — play in your browser on any device. If you like this, try Suguru Max next.