Five checks that catch the mistakes people actually make in Excel models: the one formula in a row that got edited by hand, a rate typed into a formula instead of referenced, a SUM that stops one row short. Drop a file in. It is read locally by your own browser and never sent anywhere. The page makes no network requests at all: save it and it works with the wi-fi off, which you can verify in your own developer tools.
Drop an .xlsx or .xlsm file hereor click to choose · nothing leaves this page
What it checks, and why each one matters
Inconsistent formula in a run. A row of formulas where one cell breaks the pattern. This is the commonest silent error in financial models: someone overtypes a single cell and every total downstream is quietly wrong.
Hardcoded numbers inside formulas.=B5*1.05 instead of =B5*(1+$B$6). The model looks right and cannot be re-run at a different assumption.
Error values. Cells resolving to #REF!, #DIV/0!, #VALUE!, #NAME?, #N/A, #NUM! or #NULL!.
External file references. Formulas pointing at another workbook. These break silently when the file moves and keep showing the last cached number.
Short SUM ranges. A SUM that stops one row above data that should be inside it. The classic off-by-one, and it always understates.