When should I use `int`, `float`, or `Decimal` for Python numbers?

Start with what the value means, not which type looks most familiar. Use int for whole-number values, such as a count. Use float for fractional calculations when a binary approximation is acceptable. Use Decimal when you need decimal representation or controlled rounding, as in calculations where decimal behavior matters. Python’s float uses binary floating-point, so many decimal fractions are approximations; Decimal supports exact representation for values such as Decimal('1.1').[S1][S2]

A useful first question is: “Would a small representation difference matter to this task?” If not, float may be suitable. If yes, consider Decimal and decide how your application should round results. Choosing Decimal does not automatically settle what rounding rule a task needs.

Yes. This runnable example uses 0.1 and 0.2 in both calculations, but constructs the decimal values from strings:

from decimal import Decimal

print(0.1 + 0.2)
print(Decimal("0.1") + Decimal("0.2"))

The float calculation uses binary floating-point, where those decimal fractions are approximations and operations may introduce rounding error. The Decimal calculation represents these written decimal inputs exactly and gives decimal arithmetic behavior.[S1][S2] The expressions may look similar, but their values are handled differently because they use different types.

The string matters. Decimal("0.1") constructs a decimal from the written decimal value. Decimal(0.1) instead converts the float’s exact binary value, so it does not mean the same thing as starting from the written decimal 0.1.[S1] If your input arrives as decimal text, preserve that text when constructing the Decimal.

Choose int when the value is meant to be a whole number: for example, a hypothetical count of 4 items. Don’t choose it for a quantity whose fractional part carries meaning. For a fractional calculation where an approximation is acceptable, float is the straightforward built-in floating-point option. For a task where decimal representation or controlled rounding matters, use Decimal instead.[S1][S2]

For example, if a hypothetical measurement is recorded as 2.75, float may fit a casual approximate calculation. If a hypothetical decimal amount must follow decimal arithmetic, use Decimal("2.75"). These are examples of choosing a representation, not claims about any particular application’s rules. The appropriate rounding policy depends on the task; the type alone cannot tell you what policy to use.

Construct decimal inputs from strings, and keep the calculation’s values consistently in the representation you chose. Avoid passing a float into Decimal when you intend to start from the written decimal value: converting 0.1 this way brings in the float’s exact binary value.[S1]

Also remember that Decimal arithmetic has a context controlling precision and rounding.[S1] If the result needs a specific rounding policy, determine that requirement rather than assuming the default behavior is right. The type facts here establish how these types represent values and that the context controls precision and rounding; they do not establish which rounding rule is appropriate for your particular task.