Interning is an optimization where Python reuses immutable objects (like small numbers, strings, and sometimes tuples) instead of creating new copies.

Common Examples

Strings are always interned:

a = "hello"
b = "hello"
assert a is b  # same object (string interning)

Python interns integers between -5 and 256 by default:

a = 100
b = 100
assert a is b

However, bigger numbers are not reused:

a = 1000
b = 1000
assert not a is b

Mutable and Inmutable Objects

Interning does not apply to mutable objects. Even if two lists have the same content, they will be different objects:

list1 = [1, 2, 3]
list2 = [1, 2, 3]
assert not list1 is list2

But not all immutable objects are interned either. For example, even though tuples are immutable, Python does not automatically intern them:

tuple1 = (1, 2, 3)
tuple2 = (1, 2, 3)
assert not tuple1 is tuple2

This shows that immutability allows interning, but Python only applies it to specific types like small integers and strings.

Why Does Python Do This?

Performance. It saves memory by avoiding duplicate objects.

Interning behavior can vary between implementations (CPython, PyPy…). It is just an implementation detail and may change.